Modeling dew point pressure of gas condensate reservoirs: Comparison of hybrid soft computing approaches, correlations, and thermodynamic models

Haji-Savameri, Mohammad; Menad, Nait Amar; Norouzi-Apourvari, Saeid; Hemmati-Sarapardeh, Abdolhossein

doi:10.1016/j.petrol.2019.106558

Cited by 29 publications

(6 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For small and reliable datasets, SVM framework is preferred while for larger data sets, neural network models perform better. The accuracy of predictions of each algorithm used in this study can be improved by incorporating optimization techniques such as genetic algorithm [20,21], swarm algorithm [16,28] and fuzzy logic [17].…”

Section: Resultsmentioning

confidence: 99%

“…12: End Committee Machine Intelligent System combines several models into a single model, allocating weights to each model based on their accuracy. To predict dewpoint, Haji-Savameri [28] linearly combined four ANN models into a single model using a weighted average. However, in this study, we utilized three MLP models and 1 SVM model to build the CMIS model as shown in Fig.…”

Section: Objective Functionmentioning

confidence: 99%

See 1 more Smart Citation

Performance evaluation of machine learning algorithms in predicting dew point pressure of gas condensate reservoirs

et al. 2020

View full text Add to dashboard Cite

Accurate knowledge of the dew point pressure for a gas condensate reservoir is necessary for optimizing mitigation operations during field development plan. This study explores the use of machine learning models in predicting the dew point pressure of gas condensate reservoirs. 535 experimental dew point pressure data-points with maximum temperature and pressure of 304 °F and 10,500 psi were used for this analysis. First, a standard multiple linear regression (MLR) was used as a benchmark for comparing the performance of the machine learning models. Multilayer perceptron Neural Networks (MLP) [optimized for the number of neurons and hidden layers], Support Vector Machine (SVM) [using radial basis function kernel] and Decision Tree [Gradient boost Method (GBM) and XG Boost (XGB)] algorithms were used to predict the dew point pressure. The performance of these algorithms was then compared with results obtained from published machine learning models. The input parameters for the model include; gas composition, specific gravity, the molecular weight of the heavier component and compressibility factor. The performances of these algorithms were analysed using root mean square error (RMSE), absolute average relative deviation percentage (AARD %) and coefficient of determination (R 2).

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Objective Functionmentioning

confidence: 99%

Performance evaluation of machine learning algorithms in predicting dew point pressure of gas condensate reservoirs

et al. 2020

View full text Add to dashboard Cite

show abstract

“…28 With the advancement of computation technology, these soft computing techniques have recently remerged and begun to attract attention in both academia and industries for complex system modeling. 29 Inspired from the neurological system and its basic elements for handling the information, ANNs are principally constructed as an input function x of the formal neuron i corresponding to the incoming activity (e.g., synaptic input) of the biological neuron; weight w i represents the effective magnitude of information transmission between neurons (e.g., determined by synapses), activation function z i = f(x,w i ) describes the main computation performed by a biological neuron (e.g., spike rates), and the output function y i = f(z i ) corresponds to the overall activity transmitted to the next neuron in the processing stream. 30 With the learning phase on prior representative data as a robust training and testing step, this constructed learning technique is recognized to yield good predictions in many complex systems.…”

Section: Theoretical Backgroundmentioning

confidence: 99%

Kinetic Study of Product Distribution Using Various Data-Driven and Statistical Models for Fischer–Tropsch Synthesis

Wang

Zhang

et al. 2021

ACS Omega

View full text Add to dashboard Cite

Three modeling techniques, namely, a radial basis function neural network (RBFNN), a comprehensive kinetic with genetic algorithm (CKGA), and a response surface methodology (RSM), were used to study the kinetics of Fischer−Tropsch (FT) synthesis. Using a 29 × 37 (4 independent process parameters as inputs and corresponding 36 responses as outputs) matrix with total 1073 data sets for data training through RBFNN, the established model is capable of predicting hydrocarbon product distribution i.e., the paraffin formation rate (C 2 −C 15 ) and the olefin to paraffin ratio (OPR) within acceptable uncertainties. With additional validation data sets (15 × 36 matrix with total 540 data sets), the uncertainties of using three different models were compared and the outcomes were: RBFNN (±5% uncertainties), RSM (±10% uncertainties), and CKGA (±30% uncertainties), respectively. A new effective strategy for kinetic study of the complex FT synthesis is proposed: RBFNN is used for data matrix generation with a limited number of experimental data sets (due to its fast converge and less computation time), CKGA is used for mechanism selections by the Langmuir−Hinshelwood−Hougen−Watson (LHHW) approach using a genetic algorithm to find out potential reaction pathways, and RSM is used for statistical analysis of the investigated data matrix (generated from RBFNN through central composite design) upon responses and subsequent singular/multiple optimizations. The proposed strategy is a very useful and practical tool in process engineering design and practice for the product distribution during FT synthesis.

show abstract

“…Condensate oil, as valuable light hydrocarbons with high volatility, 1 is regarded as a considerably complicated reservoir fluid due to its thermodynamic and flow properties, especially at a lower than dew point pressure. 2,3 When the pressure is below the dew point, the condensate oil near the wellbore may gradually accumulate, which results in a declined gas flow. 4−6 Since the condensate oil blockage of near-well zones can halve or even more lessen the gas well productivity, an effective evaluation of production performance of the condensate gas reservoir is required in depleted development.…”

Section: Introductionmentioning

confidence: 99%

“…Condensate oil, as valuable light hydrocarbons with high volatility, is regarded as a considerably complicated reservoir fluid due to its thermodynamic and flow properties, especially at a lower than dew point pressure. , When the pressure is below the dew point, the condensate oil near the wellbore may gradually accumulate, which results in a declined gas flow. − Since the condensate oil blockage of near-well zones can halve or even more lessen the gas well productivity, an effective evaluation of production performance of the condensate gas reservoir is required in depleted development. − The lack of sufficient knowledge about condensate gas reservoirs not only leaves huge amounts of precious condensates in the reservoir but also severely damages the gas well productivity and lowers the gas deliverability through formation of condensate oil in the reservoir, especially in near-well bore areas. − The cause of choking of condensate oil is that the heavier components in the gas are condensed isothermally because of the production-induced pressure decline, which probably results in 70–95% lowering in the relative permeability of gas. , A case study of Engineer (1985) revealed that the recovery of a Californian field was lowered to 10% of its original gas reserve due to choking of condensate oil and high saturation of moisture …”

Section: Introductionmentioning

confidence: 99%

The Application of Nuclear Magnetic Resonance 1D T₂ and 2D T₁–T₂ Maps in the Research of Condensate Oil Saturation in Condensate Gas Reservoirs

Sun

Guo

2022

Energy Fuels

View full text Add to dashboard Cite

Determination of condensate oil saturation in condensate gas reservoirs is the key parameter for estimating condensate oil blockage and gas well productivity. In PVT experiments, the constant volume depletion (CVD) method is usually adopted to obtain condensate oil saturation. However, in this method, the effect of porous media is not considered. This study applied a low field nuclear magnetic resonance (NMR) test instrument in which 1D T 2 and 2D T 1–T 2 NMR techniques are carried out to investigate the condensate oil saturation of “condensate gas-saturated” core plugs in the depletion process of 29 to 5 MPa at 25 °C. It was observed here that the T 1/T 2 ratios of condensate oil and gas at the signal peak are approximately 20 and 60, respectively, and the condensate oil and gas could be distinguished by positions in 2D NMR T 1–T 2 maps. Additionally, a standard equation (the relationship between the signal intensity and the condensate oil saturation) with an excellent linear relationship was established for the calculation of the condensate oil saturation in core plugs. The results indicated that the condensate oil first appears in micropores, and the retrograde condensation phenomenon of the core mainly comes from the contribution of mesopores and macropores; furthermore, the maximum condensate oil saturation in the porous medium is 5.36% higher than that in the PVT cylinder, and the dew point pressure in the porous medium is 11.83% higher than that in the PVT cylinder. The T 1–T 2 NMR technique may become an efficient approach for assessing fluid distributions and saturations, simultaneously possessing potential applications for characterizing fluid occurrence states in the core.

show abstract

Modeling dew point pressure of gas condensate reservoirs: Comparison of hybrid soft computing approaches, correlations, and thermodynamic models

Cited by 29 publications

References 47 publications

Performance evaluation of machine learning algorithms in predicting dew point pressure of gas condensate reservoirs

Performance evaluation of machine learning algorithms in predicting dew point pressure of gas condensate reservoirs

Kinetic Study of Product Distribution Using Various Data-Driven and Statistical Models for Fischer–Tropsch Synthesis

The Application of Nuclear Magnetic Resonance 1D T₂ and 2D T₁–T₂ Maps in the Research of Condensate Oil Saturation in Condensate Gas Reservoirs

Contact Info

Product

Resources

About

Modeling dew point pressure of gas condensate reservoirs: Comparison of hybrid soft computing approaches, correlations, and thermodynamic models

Cited by 29 publications

References 47 publications

Performance evaluation of machine learning algorithms in predicting dew point pressure of gas condensate reservoirs

Performance evaluation of machine learning algorithms in predicting dew point pressure of gas condensate reservoirs

Kinetic Study of Product Distribution Using Various Data-Driven and Statistical Models for Fischer–Tropsch Synthesis

The Application of Nuclear Magnetic Resonance 1D T2 and 2D T1–T2 Maps in the Research of Condensate Oil Saturation in Condensate Gas Reservoirs

Contact Info

Product

Resources

About

The Application of Nuclear Magnetic Resonance 1D T₂ and 2D T₁–T₂ Maps in the Research of Condensate Oil Saturation in Condensate Gas Reservoirs