abouzar choubineh - Academia.edu (original) (raw)

Papers by abouzar choubineh

Journal of Petroleum Science and Engineering, Sep 1, 2022

Journal of Computing and Information Science in Engineering, Jul 18, 2022

Deep feed-forward networks, with high complexity, backpropagate the gradient of the loss function... more Deep feed-forward networks, with high complexity, backpropagate the gradient of the loss function from final layers to earlier layers. As a consequence, the “gradient” may descend rapidly toward zero. This is known as the vanishing gradient phenomenon that prevents earlier layers from benefiting from further training. One of the most efficient techniques to solve this problem is using skip connection (shortcut) schemes that enable the gradient to be directly backpropagated to earlier layers. This paper investigates whether skip connections significantly affect the performance of deep neural networks of low complexity or whether their inclusion has little or no effect. The analysis was conducted using four Convolutional Neural Networks (CNNs) to predict four different multiscale basis functions for the mixed Generalized Multiscale Finite Element Method (GMsFEM). These models were applied to 249,375 samples. Three skip connection schemes were added to the base structure: Scheme 1 from the first convolutional block to the last, Scheme 2 from the middle to the last block, and Scheme 3 from the middle to the last and the second-to-last blocks. The results demonstrate that the third scheme is most effective, as it increases the coefficient of determination (R2) value by 0.0224–0.044 and decreases the Mean Squared Error (MSE) value by 0.0027–0.0058 compared to the base structure. Hence, it is concluded that enriching the last convolutional blocks with the information hidden in neighboring blocks is more effective than enriching using earlier convolutional blocks near the input layer.

Elsevier eBooks, 2022

Reservoir simulation methods applied to gas reservoirs are reviewed and the key influencing varia... more Reservoir simulation methods applied to gas reservoirs are reviewed and the key influencing variables identified. Machine Learning (ML) methods can be applied in various ways to improve the performance of gas reservoir simulations, especially in respect to history matching and proxy modeling. Additionally, ML can assist the CO2 sequestration and enhanced gas recovery, well placement optimization, production optimization, estimation of gas production, dew point prediction in gas condensate reservoirs, and pressure and rate transient analysis.

Journal of Petroleum Exploration and Production Technology, Aug 18, 2018

None of the various published models used to predict oil production rates through wellhead chokes... more None of the various published models used to predict oil production rates through wellhead chokes from fluid composition and pressures can be considered as a universal model for all regions. Here, a model is provided to predict liquid productionflow rates for the Reshadat oil field offshore southwest Iran, applying a customized genetic optimization algorithm (GA) and standard Excel Solver non-linear and evolutionary optimization algorithms. The dataset of 182 records of wellhead choke measurements spans liquid flow rates from < 100 to 30,000 stock tank barrels/day. Each data record includes measurements of five variables: liquid production rate (QL), wellhead pressure, choke size, basic sediment and water, and gas-liquid ratio. 70% of the dataset (127 data records) was used for training purposes to establish the prediction relationships, and 30% of the dataset (55 data records) was utilized for independently testing the accuracy of the derived relationships as predictive tools. The methodology applying either the customized GA or standard Solver optimization algorithms, demonstrates significant improvements in QL-prediction accuracy with the lowest APD (− 7.72 to − 2.89), AAPD (7.33-8.51), SD (288.77-563.85), MSE (91,871-316,429), and RMSE (303.1-562.52); and the highest R 2 (greater than 0.997) compared to six previously published liquid flow-rate prediction models. As a general result, the novel methodology is easily applied to other field/ reservoir datasets, to achieve rapid practical flow prediction applications, and is consequently of worldwide significance.

Gas Processing Journal, May 19, 2020

A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is... more A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is the interfacial tension (IFT) between formation water (brine) and injected gas. Establishing efficient and precise models for estimating CO2-brine IFT from measurements of independent variables is essential. This is the case because laboratory techniques for determining IFT are time-consuming, costly and require complex interpretation methods. For the datasets used in the current study, correlation coefficients between the input variables and measured IFT suggest that CO2 density and pressure are the most influential variables, whereas brine density is the least influential. Six artificial neural network configurations are developed and evaluated to determine their relative accuracy in predicting CO2-brine IFT. Three models involve multilayer perceptron (MLP) tuned with Levenberg-Marquardt, Bayesian regularization and scaled conjugate gradient back-propagation algorithms, respectively. Three models involve the radial basis function (RBF) trained with particle swarm optimization, differential evolution, and farmland fertility optimization algorithms, respectively. The six models all generate CO2-brine IFT predictions with high accuracy (RMSE <0.7 mN/m). However, the RBF models consistently provide slightly higher IFT prediction accuracies (RMSE <0.54 mN/m) than the MLP models.

Engineering Applications of Artificial Intelligence

Electronics

Although Deep Learning (DL) models have been introduced in various fields as effective prediction... more Although Deep Learning (DL) models have been introduced in various fields as effective prediction tools, they often do not care about uncertainty. This can be a barrier to their adoption in real-world applications. The current paper aims to apply and evaluate Monte Carlo (MC) dropout, a computationally efficient approach, to investigate the reliability of several skip connection-based Convolutional Neural Network (CNN) models while keeping their high accuracy. To do so, a high-dimensional regression problem is considered in the context of subterranean fluid flow modeling using 376,250 generated samples. The results demonstrate the effectiveness of MC dropout in terms of reliability with a Standard Deviation (SD) of 0.012–0.174, and of accuracy with a coefficient of determination (R2) of 0.7881–0.9584 and Mean Squared Error (MSE) of 0.0113–0.0508, respectively. The findings of this study may contribute to the distribution of pressure in the development of oil/gas fields.

Algorithms

Machine Learning (ML) and/or Deep Learning (DL) methods can be used to predict fluid flow in poro... more Machine Learning (ML) and/or Deep Learning (DL) methods can be used to predict fluid flow in porous media, as a suitable replacement for classical numerical approaches. Such data-driven approaches attempt to learn mappings between finite-dimensional Euclidean spaces. A novel neural framework, named Fourier Neural Operator (FNO), has been recently developed to act on infinite-dimensional spaces. A high proportion of the research available on the FNO has focused on problems with large-shape data. Furthermore, most published studies apply the FNO method to existing datasets. This paper applies and evaluates FNO to predict pressure distribution over a small, specified shape-data problem using 1700 Finite Element Method (FEM) generated samples, from heterogeneous permeability fields as the input. Considering FEM-calculated outputs as the true values, the configured FNO model provides superior prediction performance to that of a Convolutional Neural Network (CNN) in terms of statistical e...

Journal of Computing and Information Science in Engineering

As a deep feed-forward network with high complexity backpropagates the gradient of the loss funct... more As a deep feed-forward network with high complexity backpropagates the gradient of the loss function from final layers to earlier layers, the gradient might descend rapidly towards zero. This is known as the vanishing gradient phenomenon that stops the earlier layers from benefiting from further training. One of the most efficient techniques to solve this problem is using skip connection (shortcut) schemes. This paper presents an investigation of whether skip connections significantly affect the performance of deep neural networks of moderate complexity, or whether their inclusion has little or no effect. The analysis was conducted using Convolutional Neural Network (CNN) to predict four different multiscale basis functions for the mixed Generalized Multiscale Finite Element Method (GMsFEM). These models were applied to 249,375 samples generated in MatLab software, with the permeability field as the only input. Three skip connection schemes were added to the base structure: (Scheme ...

Journal of Petroleum Science and Engineering

Petroleum Science and Technology, 2016

ABSTRACT The authors' aim was to present new models based on artificial neural network (ANN) ... more ABSTRACT The authors' aim was to present new models based on artificial neural network (ANN) and two optimization algorithms including cuckoo optimization algorithm (COA) and teaching learning based optimization (TLBO) to predict the pure and impure CO2 MMP. Thirty-four and 11 training and testing data sets were used to develop these models with following inputs: reservoir temperature, the mole percent of volatile oil components (C1 and N2), mole percent of intermediate oil components (C2-C4, CO2, and H2S), molecular weight of C5+ fraction in oil phase (MWC5+) and mole percentage of CO2, N2, C1, C4, and H2S in the injected gas. Statistical comparisons show that although two models yield acceptable results, the ANN-TLBO model has better performance with the lower mean absolute percentage error (2.6%) and standard deviation (3.37%) and the higher coefficient of determination (0.993). Moreover, among the available correlations, the Cronquist's (1978; corrected by Sebastian et al., 1985) correlations have better performance. Finally, the sensitivity analysis on the ANN-TLBO showed that MWC5+ and reservoir temperature are the most influential parameters in determining the CO2 MMP, respectively.

Petroleum Science and Technology, 2017

ABSTRACT Proper calculations of gas engineering require precise determination of gas properties a... more ABSTRACT Proper calculations of gas engineering require precise determination of gas properties and its associated variations with pressure and temperature. These properties can be determined by conducting experimental tests on gathered fluid samples from the bottom of the wellbore or at the surface as well as using equations of state and empirical correlations. This work is concentrated to develop a robust and quick model based on artificial network trained with teaching learning based optimization (ANN-TLBO) using 693 data sets at a wide range of pressure and temperature for gas density prediction. Comparing gas density from the predictive method and experimental results describe that the proposed ANN-TLBO model is of reliable accuracy for determining gas density. Sensitivity analysis also showed the extreme effect of temperature and pressure on gas density.

Machine-learning algorithms aid predictions for complex systems with multiple influencing variabl... more Machine-learning algorithms aid predictions for complex systems with multiple influencing variables. However, many neural-network related algorithms behave as black boxes in terms of revealing how the prediction of each data record is performed. This drawback limits their ability to provide detailed insights concerning the workings of the underlying system, or to relate predictions to specific characteristics of the underlying variables. The recently proposed transparent open box (TOB) learning network algorithm successfully addresses these issues by revealing the exact calculation involved in the prediction of each data record. That algorithm, described in summary, can be applied in a spreadsheet or fully-coded configurations and offers significant benefits to analysis and prediction of many natural gas systems. The algorithm is applied to the prediction of natural gas density using a published dataset of 693 data records involving 14 variables (temperature and pressure plus the m...

Natural gas reservoir simulation, as a physics-based numerical method, needs to be carried out wi... more Natural gas reservoir simulation, as a physics-based numerical method, needs to be carried out with a high level of precision. If not, it may be highly misleading and cause substantial losses, poor estimation of ultimate recovery factor, and wasted effort. Although simple simulations often provide acceptable approximations, there is a continued desire to develop more sophisticated simulation strategies and techniques. Given the capabilities of Machine Learning (ML) and their general acceptance in recent decades, this chapter considers the application of these techniques to gas reservoir simulations. The aspiration ML technics should be capable of providing some improvements in terms of both accuracy and speed. The simulation of gas reservoirs (dry gas, wet gas and retrograde gas-condensate) is introduced along with its fundamental concepts and governing equations. More specific and advanced concepts of applying ML in modern reservoir simulation models are described and justified, pa...

A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is... more A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is the interfacial tension (IFT) between formation water (brine) and injected gas. Establishing efficient and precise models for estimating CO2 – brine IFT from measurements of independent variables is essential. This is the case, because laboratory techniques for determining IFT are time-consuming, costly and require complex interpretation methods. For the datasets used in the current study, correlation coefficients between the input variables and measured IFT suggests that CO2 density and pressure are the most influential variables, whereas brine density is the least influential. Six artificial neural network configurations are developed and evaluated to determine their relative accuracy in predicting CO2 – brine IFT. Three models involve multilayer perceptron (MLP) tuned with Levenberg-Marquardt, Bayesian regularization and scaled conjugate gradient back-propagation algorithms, respecti...

International Journal of Critical Infrastructure Protection

Journal of Natural Gas Geoscience

Experimental and Computational Multiphase Flow

Pressure-Volume-Temperature (PVT) characterization of a crude oil involves establishing its bubbl... more Pressure-Volume-Temperature (PVT) characterization of a crude oil involves establishing its bubble point pressure, which is the pressure at which the first gas bubble forms on a fluid sample while reducing pressure at a stabilized temperature. Although accurate measurement can be made experimentally, such experiments are expensive and time-consuming. Consequently, applying reliable artificial intelligence (AI)/machine learning methods to provide an accurate mathematical prediction of an oil's bubble point pressure from more easily measured characteristics can provide valuable cost and time savings. This paper develops and compares four neurocomputing models applying algorithms consisting of a Multilayer Perceptron (MLP), a Radial Basis Function trained with a Genetic Algorithm (RBF-GA), a Combined Hybrid Particle Swarm Optimization-Adaptive Neuro-Fuzzy Inference System (CHPSO-ANFIS), and Least Squared Support Vector Machine (LSSVM) tuned with a coupled simulated annealing (CSA) optimizer. Based on a comprehensive analysis, although the four proposed models yield acceptable outputs, the CHPSO-ANFIS model has the best performance with the average absolute relative deviation of 0.846, the standard deviation of 0.0126, the root mean square error of 43.21, and the correlation coefficient of 0.9902. These algorithms are deployed for the accurate estimation of the bubble point pressure from the giant Ahvaz oil field (Iran).

Journal of Petroleum Exploration and Production Technology

None of the various published models used to predict oil production rates through wellhead chokes... more None of the various published models used to predict oil production rates through wellhead chokes from fluid composition and pressures can be considered as a universal model for all regions. Here, a model is provided to predict liquid productionflow rates for the Reshadat oil field offshore southwest Iran, applying a customized genetic optimization algorithm (GA) and standard Excel Solver non-linear and evolutionary optimization algorithms. The dataset of 182 records of wellhead choke measurements spans liquid flow rates from < 100 to 30,000 stock tank barrels/day. Each data record includes measurements of five variables: liquid production rate (QL), wellhead pressure, choke size, basic sediment and water, and gas-liquid ratio. 70% of the dataset (127 data records) was used for training purposes to establish the prediction relationships, and 30% of the dataset (55 data records) was utilized for independently testing the accuracy of the derived relationships as predictive tools. The methodology applying either the customized GA or standard Solver optimization algorithms, demonstrates significant improvements in QL-prediction accuracy with the lowest APD (− 7.72 to − 2.89), AAPD (7.33-8.51), SD (288.77-563.85), MSE (91,871-316,429), and RMSE (303.1-562.52); and the highest R 2 (greater than 0.997) compared to six previously published liquid flow-rate prediction models. As a general result, the novel methodology is easily applied to other field/ reservoir datasets, to achieve rapid practical flow prediction applications, and is consequently of worldwide significance.

Advances in Geo-Energy Research

Neural-network, machine-learning algorithms are effective prediction tools but can behave as blac... more Neural-network, machine-learning algorithms are effective prediction tools but can behave as black boxes in many applications by not easily providing the exact calculations and relationships among the underlying input variables (which may or may not be independent of each other) involved each of their predictions. The transparent open box (TOB) learning network algorithm overcomes this limitation by providing the exact calculations involved in all its predictions and achieving acceptable and auditable levels of prediction accuracy. The TOB network, based on an optimized data-matching algorithm, can be applied in spreadsheet or fully-coded configurations. This algorithm offers significant benefits to analysis and prediction of many complex and difficult to measure non-linear systems. To demonstrate its prediction performance, the algorithm is applied to the prediction of crude oil formation volume factor at bubble point (B ob) using published datasets of 166, 203 and 237 data records involving 4 variables (reservoir temperature, gas-oil ratio, oil gravity and gas specific gravity). Two of these datasets display uneven and irregular data coverage. The TOB network demonstrates high prediction accuracy for B ob (Root Mean Square Error (RMSE) ∼ 0.03; R 2 > 0.95) for the more evenly distributed dataset. The performance of the TOB readily reveals the risk of overfitting such datasets. With its high levels of transparency and inhibitions to being overfitted, the TOB learning network offers an insightful approach to machine learning applied to predicting complex nonlinear systems. Its results complement and benchmark the prediction contributions of neural networks and empirical correlations. In doing so it provides further insight to the underlying data.

Journal of Petroleum Science and Engineering, Sep 1, 2022

Journal of Computing and Information Science in Engineering, Jul 18, 2022

Deep feed-forward networks, with high complexity, backpropagate the gradient of the loss function... more Deep feed-forward networks, with high complexity, backpropagate the gradient of the loss function from final layers to earlier layers. As a consequence, the “gradient” may descend rapidly toward zero. This is known as the vanishing gradient phenomenon that prevents earlier layers from benefiting from further training. One of the most efficient techniques to solve this problem is using skip connection (shortcut) schemes that enable the gradient to be directly backpropagated to earlier layers. This paper investigates whether skip connections significantly affect the performance of deep neural networks of low complexity or whether their inclusion has little or no effect. The analysis was conducted using four Convolutional Neural Networks (CNNs) to predict four different multiscale basis functions for the mixed Generalized Multiscale Finite Element Method (GMsFEM). These models were applied to 249,375 samples. Three skip connection schemes were added to the base structure: Scheme 1 from the first convolutional block to the last, Scheme 2 from the middle to the last block, and Scheme 3 from the middle to the last and the second-to-last blocks. The results demonstrate that the third scheme is most effective, as it increases the coefficient of determination (R2) value by 0.0224–0.044 and decreases the Mean Squared Error (MSE) value by 0.0027–0.0058 compared to the base structure. Hence, it is concluded that enriching the last convolutional blocks with the information hidden in neighboring blocks is more effective than enriching using earlier convolutional blocks near the input layer.

Elsevier eBooks, 2022

Reservoir simulation methods applied to gas reservoirs are reviewed and the key influencing varia... more Reservoir simulation methods applied to gas reservoirs are reviewed and the key influencing variables identified. Machine Learning (ML) methods can be applied in various ways to improve the performance of gas reservoir simulations, especially in respect to history matching and proxy modeling. Additionally, ML can assist the CO2 sequestration and enhanced gas recovery, well placement optimization, production optimization, estimation of gas production, dew point prediction in gas condensate reservoirs, and pressure and rate transient analysis.

Journal of Petroleum Exploration and Production Technology, Aug 18, 2018

None of the various published models used to predict oil production rates through wellhead chokes... more None of the various published models used to predict oil production rates through wellhead chokes from fluid composition and pressures can be considered as a universal model for all regions. Here, a model is provided to predict liquid productionflow rates for the Reshadat oil field offshore southwest Iran, applying a customized genetic optimization algorithm (GA) and standard Excel Solver non-linear and evolutionary optimization algorithms. The dataset of 182 records of wellhead choke measurements spans liquid flow rates from < 100 to 30,000 stock tank barrels/day. Each data record includes measurements of five variables: liquid production rate (QL), wellhead pressure, choke size, basic sediment and water, and gas-liquid ratio. 70% of the dataset (127 data records) was used for training purposes to establish the prediction relationships, and 30% of the dataset (55 data records) was utilized for independently testing the accuracy of the derived relationships as predictive tools. The methodology applying either the customized GA or standard Solver optimization algorithms, demonstrates significant improvements in QL-prediction accuracy with the lowest APD (− 7.72 to − 2.89), AAPD (7.33-8.51), SD (288.77-563.85), MSE (91,871-316,429), and RMSE (303.1-562.52); and the highest R 2 (greater than 0.997) compared to six previously published liquid flow-rate prediction models. As a general result, the novel methodology is easily applied to other field/ reservoir datasets, to achieve rapid practical flow prediction applications, and is consequently of worldwide significance.

Gas Processing Journal, May 19, 2020

A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is... more A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is the interfacial tension (IFT) between formation water (brine) and injected gas. Establishing efficient and precise models for estimating CO2-brine IFT from measurements of independent variables is essential. This is the case because laboratory techniques for determining IFT are time-consuming, costly and require complex interpretation methods. For the datasets used in the current study, correlation coefficients between the input variables and measured IFT suggest that CO2 density and pressure are the most influential variables, whereas brine density is the least influential. Six artificial neural network configurations are developed and evaluated to determine their relative accuracy in predicting CO2-brine IFT. Three models involve multilayer perceptron (MLP) tuned with Levenberg-Marquardt, Bayesian regularization and scaled conjugate gradient back-propagation algorithms, respectively. Three models involve the radial basis function (RBF) trained with particle swarm optimization, differential evolution, and farmland fertility optimization algorithms, respectively. The six models all generate CO2-brine IFT predictions with high accuracy (RMSE <0.7 mN/m). However, the RBF models consistently provide slightly higher IFT prediction accuracies (RMSE <0.54 mN/m) than the MLP models.

Engineering Applications of Artificial Intelligence

Electronics

Although Deep Learning (DL) models have been introduced in various fields as effective prediction... more Although Deep Learning (DL) models have been introduced in various fields as effective prediction tools, they often do not care about uncertainty. This can be a barrier to their adoption in real-world applications. The current paper aims to apply and evaluate Monte Carlo (MC) dropout, a computationally efficient approach, to investigate the reliability of several skip connection-based Convolutional Neural Network (CNN) models while keeping their high accuracy. To do so, a high-dimensional regression problem is considered in the context of subterranean fluid flow modeling using 376,250 generated samples. The results demonstrate the effectiveness of MC dropout in terms of reliability with a Standard Deviation (SD) of 0.012–0.174, and of accuracy with a coefficient of determination (R2) of 0.7881–0.9584 and Mean Squared Error (MSE) of 0.0113–0.0508, respectively. The findings of this study may contribute to the distribution of pressure in the development of oil/gas fields.

Algorithms

Machine Learning (ML) and/or Deep Learning (DL) methods can be used to predict fluid flow in poro... more Machine Learning (ML) and/or Deep Learning (DL) methods can be used to predict fluid flow in porous media, as a suitable replacement for classical numerical approaches. Such data-driven approaches attempt to learn mappings between finite-dimensional Euclidean spaces. A novel neural framework, named Fourier Neural Operator (FNO), has been recently developed to act on infinite-dimensional spaces. A high proportion of the research available on the FNO has focused on problems with large-shape data. Furthermore, most published studies apply the FNO method to existing datasets. This paper applies and evaluates FNO to predict pressure distribution over a small, specified shape-data problem using 1700 Finite Element Method (FEM) generated samples, from heterogeneous permeability fields as the input. Considering FEM-calculated outputs as the true values, the configured FNO model provides superior prediction performance to that of a Convolutional Neural Network (CNN) in terms of statistical e...

Journal of Computing and Information Science in Engineering

As a deep feed-forward network with high complexity backpropagates the gradient of the loss funct... more As a deep feed-forward network with high complexity backpropagates the gradient of the loss function from final layers to earlier layers, the gradient might descend rapidly towards zero. This is known as the vanishing gradient phenomenon that stops the earlier layers from benefiting from further training. One of the most efficient techniques to solve this problem is using skip connection (shortcut) schemes. This paper presents an investigation of whether skip connections significantly affect the performance of deep neural networks of moderate complexity, or whether their inclusion has little or no effect. The analysis was conducted using Convolutional Neural Network (CNN) to predict four different multiscale basis functions for the mixed Generalized Multiscale Finite Element Method (GMsFEM). These models were applied to 249,375 samples generated in MatLab software, with the permeability field as the only input. Three skip connection schemes were added to the base structure: (Scheme ...

Journal of Petroleum Science and Engineering

Petroleum Science and Technology, 2016

ABSTRACT The authors' aim was to present new models based on artificial neural network (ANN) ... more ABSTRACT The authors' aim was to present new models based on artificial neural network (ANN) and two optimization algorithms including cuckoo optimization algorithm (COA) and teaching learning based optimization (TLBO) to predict the pure and impure CO2 MMP. Thirty-four and 11 training and testing data sets were used to develop these models with following inputs: reservoir temperature, the mole percent of volatile oil components (C1 and N2), mole percent of intermediate oil components (C2-C4, CO2, and H2S), molecular weight of C5+ fraction in oil phase (MWC5+) and mole percentage of CO2, N2, C1, C4, and H2S in the injected gas. Statistical comparisons show that although two models yield acceptable results, the ANN-TLBO model has better performance with the lower mean absolute percentage error (2.6%) and standard deviation (3.37%) and the higher coefficient of determination (0.993). Moreover, among the available correlations, the Cronquist's (1978; corrected by Sebastian et al., 1985) correlations have better performance. Finally, the sensitivity analysis on the ANN-TLBO showed that MWC5+ and reservoir temperature are the most influential parameters in determining the CO2 MMP, respectively.

Petroleum Science and Technology, 2017

ABSTRACT Proper calculations of gas engineering require precise determination of gas properties a... more ABSTRACT Proper calculations of gas engineering require precise determination of gas properties and its associated variations with pressure and temperature. These properties can be determined by conducting experimental tests on gathered fluid samples from the bottom of the wellbore or at the surface as well as using equations of state and empirical correlations. This work is concentrated to develop a robust and quick model based on artificial network trained with teaching learning based optimization (ANN-TLBO) using 693 data sets at a wide range of pressure and temperature for gas density prediction. Comparing gas density from the predictive method and experimental results describe that the proposed ANN-TLBO model is of reliable accuracy for determining gas density. Sensitivity analysis also showed the extreme effect of temperature and pressure on gas density.

Machine-learning algorithms aid predictions for complex systems with multiple influencing variabl... more Machine-learning algorithms aid predictions for complex systems with multiple influencing variables. However, many neural-network related algorithms behave as black boxes in terms of revealing how the prediction of each data record is performed. This drawback limits their ability to provide detailed insights concerning the workings of the underlying system, or to relate predictions to specific characteristics of the underlying variables. The recently proposed transparent open box (TOB) learning network algorithm successfully addresses these issues by revealing the exact calculation involved in the prediction of each data record. That algorithm, described in summary, can be applied in a spreadsheet or fully-coded configurations and offers significant benefits to analysis and prediction of many natural gas systems. The algorithm is applied to the prediction of natural gas density using a published dataset of 693 data records involving 14 variables (temperature and pressure plus the m...

Natural gas reservoir simulation, as a physics-based numerical method, needs to be carried out wi... more Natural gas reservoir simulation, as a physics-based numerical method, needs to be carried out with a high level of precision. If not, it may be highly misleading and cause substantial losses, poor estimation of ultimate recovery factor, and wasted effort. Although simple simulations often provide acceptable approximations, there is a continued desire to develop more sophisticated simulation strategies and techniques. Given the capabilities of Machine Learning (ML) and their general acceptance in recent decades, this chapter considers the application of these techniques to gas reservoir simulations. The aspiration ML technics should be capable of providing some improvements in terms of both accuracy and speed. The simulation of gas reservoirs (dry gas, wet gas and retrograde gas-condensate) is introduced along with its fundamental concepts and governing equations. More specific and advanced concepts of applying ML in modern reservoir simulation models are described and justified, pa...

A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is... more A key variable for determining carbon dioxide (CO2) storage capacity in sub-surface reservoirs is the interfacial tension (IFT) between formation water (brine) and injected gas. Establishing efficient and precise models for estimating CO2 – brine IFT from measurements of independent variables is essential. This is the case, because laboratory techniques for determining IFT are time-consuming, costly and require complex interpretation methods. For the datasets used in the current study, correlation coefficients between the input variables and measured IFT suggests that CO2 density and pressure are the most influential variables, whereas brine density is the least influential. Six artificial neural network configurations are developed and evaluated to determine their relative accuracy in predicting CO2 – brine IFT. Three models involve multilayer perceptron (MLP) tuned with Levenberg-Marquardt, Bayesian regularization and scaled conjugate gradient back-propagation algorithms, respecti...

International Journal of Critical Infrastructure Protection

Journal of Natural Gas Geoscience

Experimental and Computational Multiphase Flow

Pressure-Volume-Temperature (PVT) characterization of a crude oil involves establishing its bubbl... more Pressure-Volume-Temperature (PVT) characterization of a crude oil involves establishing its bubble point pressure, which is the pressure at which the first gas bubble forms on a fluid sample while reducing pressure at a stabilized temperature. Although accurate measurement can be made experimentally, such experiments are expensive and time-consuming. Consequently, applying reliable artificial intelligence (AI)/machine learning methods to provide an accurate mathematical prediction of an oil's bubble point pressure from more easily measured characteristics can provide valuable cost and time savings. This paper develops and compares four neurocomputing models applying algorithms consisting of a Multilayer Perceptron (MLP), a Radial Basis Function trained with a Genetic Algorithm (RBF-GA), a Combined Hybrid Particle Swarm Optimization-Adaptive Neuro-Fuzzy Inference System (CHPSO-ANFIS), and Least Squared Support Vector Machine (LSSVM) tuned with a coupled simulated annealing (CSA) optimizer. Based on a comprehensive analysis, although the four proposed models yield acceptable outputs, the CHPSO-ANFIS model has the best performance with the average absolute relative deviation of 0.846, the standard deviation of 0.0126, the root mean square error of 43.21, and the correlation coefficient of 0.9902. These algorithms are deployed for the accurate estimation of the bubble point pressure from the giant Ahvaz oil field (Iran).

Journal of Petroleum Exploration and Production Technology

None of the various published models used to predict oil production rates through wellhead chokes... more None of the various published models used to predict oil production rates through wellhead chokes from fluid composition and pressures can be considered as a universal model for all regions. Here, a model is provided to predict liquid productionflow rates for the Reshadat oil field offshore southwest Iran, applying a customized genetic optimization algorithm (GA) and standard Excel Solver non-linear and evolutionary optimization algorithms. The dataset of 182 records of wellhead choke measurements spans liquid flow rates from < 100 to 30,000 stock tank barrels/day. Each data record includes measurements of five variables: liquid production rate (QL), wellhead pressure, choke size, basic sediment and water, and gas-liquid ratio. 70% of the dataset (127 data records) was used for training purposes to establish the prediction relationships, and 30% of the dataset (55 data records) was utilized for independently testing the accuracy of the derived relationships as predictive tools. The methodology applying either the customized GA or standard Solver optimization algorithms, demonstrates significant improvements in QL-prediction accuracy with the lowest APD (− 7.72 to − 2.89), AAPD (7.33-8.51), SD (288.77-563.85), MSE (91,871-316,429), and RMSE (303.1-562.52); and the highest R 2 (greater than 0.997) compared to six previously published liquid flow-rate prediction models. As a general result, the novel methodology is easily applied to other field/ reservoir datasets, to achieve rapid practical flow prediction applications, and is consequently of worldwide significance.

Advances in Geo-Energy Research

Neural-network, machine-learning algorithms are effective prediction tools but can behave as blac... more Neural-network, machine-learning algorithms are effective prediction tools but can behave as black boxes in many applications by not easily providing the exact calculations and relationships among the underlying input variables (which may or may not be independent of each other) involved each of their predictions. The transparent open box (TOB) learning network algorithm overcomes this limitation by providing the exact calculations involved in all its predictions and achieving acceptable and auditable levels of prediction accuracy. The TOB network, based on an optimized data-matching algorithm, can be applied in spreadsheet or fully-coded configurations. This algorithm offers significant benefits to analysis and prediction of many complex and difficult to measure non-linear systems. To demonstrate its prediction performance, the algorithm is applied to the prediction of crude oil formation volume factor at bubble point (B ob) using published datasets of 166, 203 and 237 data records involving 4 variables (reservoir temperature, gas-oil ratio, oil gravity and gas specific gravity). Two of these datasets display uneven and irregular data coverage. The TOB network demonstrates high prediction accuracy for B ob (Root Mean Square Error (RMSE) ∼ 0.03; R 2 > 0.95) for the more evenly distributed dataset. The performance of the TOB readily reveals the risk of overfitting such datasets. With its high levels of transparency and inhibitions to being overfitted, the TOB learning network offers an insightful approach to machine learning applied to predicting complex nonlinear systems. Its results complement and benchmark the prediction contributions of neural networks and empirical correlations. In doing so it provides further insight to the underlying data.