Explicit Neural Network-Based Models for Bubble Point Pressure and Formation Volume Factor Prediction

Okon, Anietie N.; Effiong, Augustine James; Daniel, Deborah David

doi:10.1007/s13369-022-07240-3

Cited by 5 publications

(6 citation statements)

References 95 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is because the R and MSE values obtained for the various networks are closer to 1, with their respective MSE values closer to zero. From a statistical standpoint, the R and MSE values obtained for the networks trained with clip-normalized datasets strongly indicate the networks' good prediction performance[25,28].…”

mentioning

confidence: 88%

“…where 𝑦 𝑠𝑐𝑎𝑙𝑒𝑑 denotes the scaled values for input or output parameters, 𝑦 𝑖 is the values of the non-normalized parameters, 𝑦 𝑚𝑖𝑛 and 𝑦 𝑚𝑎𝑥 represent the minimum and maximum values of the non-normalized parameters, respectively. According to Okon and Ansa [24] and Okon et al [25], normalizing the datasets for the neural network training is necessary for the following reasons: adequate adjustment of the network connecting weights for optimum prediction and reducing the sensitivity of the sigmoidal (i.e. transfer or activation) function to large datasets values.…”

Section: Data Acquisition and Preparationmentioning

confidence: 99%

“…integral values) were provided to the network as output data. In the training phase, the network learning algorithm Levenberg-Marquardt (trainlm in MATLAB) adjusts the weights and biases based on the feedforward back-propagation (FFBP) method until the neurons learn the input-output data structure [25]. The network learning algorithm (Levenberg-Marquardt) was chosen because it is faster (i.e., takes less training time) and suitable for non-linear functions minimization compared to other learning algorithms, namely Bayesian regularization and scaled conjugate gradient, available in the MATLAB software.…”

Section: Neural Network Development and Trainingmentioning

confidence: 99%

“…The chosen activation functions are to ensure conformity with the maximumminimum normalization approach output values range, that is, 0 to 1. According to Okon et al [25], the weights and biases that produce the lowest error from the supervised datasets are the best generalizations of the network. Thus, the lowest MSE and the correlation coefficient (R) values were also used to select the best network topology (architecture).…”

Section: Neural Network Development and Trainingmentioning

confidence: 99%

“…After several trials of different network topologies, the best network performance was obtained with three topologies: 2-3-1, 2-4-1, and 2-5-1. For more information on the neural network stopping criteria, weights generation and adjustment, neural network computations, and generalisation, readers can read published works by Mahmoudi and Mahmoudi [26], Okon and Ansa [24], and Okon et al [25]. As a result, Table 3…”

Section: Neural Network Development and Trainingmentioning

confidence: 99%

See 4 more Smart Citations

Interpretation of Neural Network Models – New Insights of Estimating Static Bottom-Hole Pressures of Gas Wells in Nigerian Petroleum Provinces

Udoma,

Okon,

Udoeyop

et al. 2023

Int. J. Petrol. Technol.

View full text Add to dashboard Cite

Methods to determine static bottom-hole pressure (BHP) from surface measurements include the average temperature and z-factor method, the Sukkar-Cornell method, the Cullender-Smith method, and the Poettmann method. Among these methods, the Poettmann method is preferable in the petroleum industry but with a concern for software developers, as the integral values to determine the static BHP are tabular. In this study, neural network-based models to predict the integral values using pseudo-reduced pressures and temperatures were developed. The 2-3-1, 2-4-1, and 2-5-1 neural-based models had overall correlation coefficients (R) of 0.9974, 0.99835, and 0.99745, respectively, for the maximum-minimum normalization method and R of 0.99745, 0.99805, and 0.9992 for the clip-scaling method. Comparing the models' predictions with the Lagrangian interpolated values resulted in R of 0.99895 and 0.9995 for the maximum-minimum and clip-scaling-based models. Thus, the developed models can predict Poettmann's integral values without table look-up to estimate static BHP in gas wells.

show abstract

mentioning

confidence: 88%

Section: Data Acquisition and Preparationmentioning

confidence: 99%

Section: Neural Network Development and Trainingmentioning

confidence: 99%

Section: Neural Network Development and Trainingmentioning

confidence: 99%

Section: Neural Network Development and Trainingmentioning

confidence: 99%

See 3 more Smart Citations

Interpretation of Neural Network Models – New Insights of Estimating Static Bottom-Hole Pressures of Gas Wells in Nigerian Petroleum Provinces

Udoma,

Okon,

Udoeyop

et al. 2023

Int. J. Petrol. Technol.

View full text Add to dashboard Cite

show abstract

Ion-adsorbed REE clays: Swelling challenges and future solutions

Negash,

Effiong,

Khan

et al. 2024

Journal of Molecular Liquids

View full text Add to dashboard Cite

Simplified Neural Network-Based Models for Oil Flow Rate Prediction

Umana,

Okon,

Agwu

2024

PSE

View full text Add to dashboard Cite

Available neural network-based models for predicting the oil flow rate (qo) in the Niger Delta are not simplified and are developed from limited data sources. The reproducibility of these models is not feasible as the models’ details are not published. This study developed simplified and reproducible three, five, and six-input variables neural-based models for estimating qo using 283 datasets from 21 wells across fields in the Niger Delta. The neural-based models were developed using maximum-minimum (max.-min.) normalized and clip-normalized datasets. The performances and the generalizability of the developed models with published datasets were determined using some statistical indices: coefficient of determination (R2), mean square error (MSE), root mean square error (RMSE), average relative error (ARE) and average absolute relative error (AARE). The results indicate that the 3-input-based neural models had overall R2, MSE, and RMSE values of 0.9689, 9.6185x10-4 and 0.0310, respectively, for the max.-min. normalizing method and R2 of 0.9663, MSE of 5.7986x10-3 and RMSE of 0.0762 for the clip scaling approach. The 5-input-based models resulted in R2 of 0.9865, MSE of 5.7790×10-4 and RMSE of 0.0240 for the max.-min. scaling method and R2 of 0.9720, MSE of 3.7243x10-3 and RMSE of 0.0610 for the clip scaling approach. Also, the 6-input-based models had R2 of 0.9809, MSE of 8.7520x10-4 and RMSE of 0.0296 for the max.-min. normalizing approach and R2 of 0.9791, MSE of 3.8859 x 10-3 and RMSE of 0.0623 for the clip scaling method. Furthermore, the generality performance of the simplified neural-based models resulted in R2, RMSE, ARE, and AAPRE of 0.9644, 205.78, 0.0248, and 0.1275, respectively, for the 3-input-based neural model and R2 of 0.9264, RMSE of 2089.93, ARE of 0.1656 and AARE of 0.2267 for the 6-input-based neural model. The neural-based models predicted qo were more comparable to the test datasets than some existing correlations, as the predicted qo result was the lowest error indices. Besides, the overall relative importance of the neural-based models’ input variables on qo prediction is S>GLR>Pwh>T/Tsc>γo>BS&W>γg. The simplified neural-based models performed better than some empirical correlations from the assessment indicators. Therefore, the models should apply as tools for oil flow rate prediction in the Niger Delta fields, as the necessary details to implement the models are made visible.

show abstract

Explicit Neural Network-Based Models for Bubble Point Pressure and Formation Volume Factor Prediction

Cited by 5 publications

References 95 publications

Interpretation of Neural Network Models – New Insights of Estimating Static Bottom-Hole Pressures of Gas Wells in Nigerian Petroleum Provinces

Interpretation of Neural Network Models – New Insights of Estimating Static Bottom-Hole Pressures of Gas Wells in Nigerian Petroleum Provinces

Ion-adsorbed REE clays: Swelling challenges and future solutions

Simplified Neural Network-Based Models for Oil Flow Rate Prediction

Contact Info

Product

Resources

About