The Application of an Artificial Neural Network (ANN) and a Genetic Programming Neural Network (GPNN) for the Modeling of Experimental Data of Slim Tube Permeability Reduction by Asphaltene Precipitation in Iranian Crude Oil Reservoirs

Abstract

The authors introduce a new implementation of the neural network (ANN), genetic programming neural network (GPNN), and neuro-fuzzy (NF) technology in petroleum engineering. An intelligent framework is developed for calculating the amount of permeability reduction by asphaltene precipitation in Iranian crude oil reservoirs over a wide pressure, temperature, and solvent mole fraction range. Theoretical results and practical experience indicate that a feed-forward network can approximate a wide class of function relationships very well. In this work, a conventional feed-forward multilayer ANN, GPNN, and NF approach have been proposed to predict the amount of permeability reduction. The accuracy of the method is evaluated by predicting the amount of permeability reduction of various reservoir fluids not used in the development of the models. One of the ways in modeling such systems is using intelligent techniques, which need information about the systems, so, based on some intelligent learning methods, it can provide a suitable model. Furthermore, the performance of the model is compared with the performance of a simple model for permeability reduction prediction, a new correlation, and experimental data. Results of this comparison show that the proposed GPNN method first and then NF method is superior both in accuracy and generality, over the other models.     

Intelligent Modeling of Asphaltene Precipitation in Live and Tank Crude Oil Systems

Abstract

The study of asphaltene precipitation properties has been motivated by their propensity to aggregate, flocculate, precipitate, and adsorb onto interfaces. The tendency of asphaltenes to precipitation has posed great challenges for the petroleum industry. The most important parameters in asphaltene precipitation modeling and prediction are the asphaltene and oil solvent solubility parameters, which are very sensitive to reservoir and operational conditions. The driving force of asphaltene flocculation is the difference between asphaltene and the oil solvent solubility parameter. Since the nature of asphaltene solubility is yet unknown and several unmodeled dynamics are hidden in the original systems, the existing prediction models may fail in prediction the asphaltene precipitation in crude oil systems. One of ways in modeling such systems is using intelligent techniques that need some information about the systems; so, based on some intelligent learning methods it can provide a suitable model. The authors introduce a new implementation of the artificial intelligent computing technology in petroleum engineering. They have proposed a new approach to prediction of the asphaltene precipitation in crude oil systems using fuzzy logic, neural networks, and genetic algorithms. Results of this research indicate that the proposed prediction model with recognizing the possible patterns between input and output variables can successfully predict and model asphaltene precipitation in tank and live crude oils with a good accuracy.     

Modeling Asphaltene Precipitation Under Gas Injection and Pressure Depletion Conditions

Asphaltene precipitation is a major problem during primary oil production and enhanced oil recovery in the petroleum industry. In this work, a series of experiments was carried to determine the asphaltene precipitation of bottom hole live oil during gas injection and pressure depletion condition with Iranian bottom hole live oil sample, which is close to reservoir conditions using high pressure-high temperature equilibrium cell. In the majority of previous works, the mixture of recombined oil (mixture dead oil and associated gas) was used which is far from reservoir conditions. The used pressure ranges in this work covers wide ranges from 3 to 35 MPa for natural depletion processes and 24–45 MPa for gas injection processes. Also, a new approach based on the artificial neural network (ANN) method has been developed to account the asphaltene precipitation under pressure depletion/gas injection conditions and the proposed model was verified using experimental data reported in the literature and in this work. A three-layer feed-forward ANN by using the Levenberg-Marquardt back-propagation optimization algorithm for network training has been used in proposed artificial neural network model. The maximum mean square error of 0.001191 has been found. In order to compare the performance of the proposed model based on artificial neural network method, the asphaltene precipitation experimental data under pressure depletion/gas injection conditions were correlated using Solid and Flory-Huggins models. The results show that the proposed model based on artificial neural network method predicts more accurately the asphaltene precipitation experimental data in comparison to other models with deviation of less than 5%. Also, the number of parameters required for the ANN model is less than the studied thermodynamic models. It should be noted that the Flory and solid models can correlate accurately the asphaltene precipitation during methane injection in comparison with CO₂ injection.     

Phase behavior modeling of asphaltene precipitation utilizing RBF-ANN approach

Precipitation of heavy hydrocarbons, particularly asphaltenes, is the reason for numerous operational and production problems in the petroleum industry. Hence, knowing the amount of asphaltene precipitation is a critical commission for petroleum engineers to overcome its problems. The aim of this study was to predict the amount of asphaltene precipitation as a function of temperature, dilution ratio, and molecular weight of different n-alkanes utilizing radial basis function artificial neural network (RBF-ANN). Additionally, this model has been compared with previous correlations, and its great accuracy was proved to predict the precipitated asphaltene. The values of R-squared and mean squared error obtained were 0.998 and 0.007, respectively. The efforts confirmed brilliant forecasting skill of RBF-ANN for the approximation of the precipitated asphaltene as a function of temperature, dilution ratio, and molecular weight of different n-alkanes.     

PREDICTING THE EFFECT OF FEEDSTOCK ON PRODUCT YIELDS AND PROPERTIES OF THE FCC PROCESS

ABSTRACT

The mechanism of petroleum refining processes are too complex, and no thorough model has yet been developed. Neural networks represent an effective alternative to mathematical modeling of refinery operations if a sufficient amount of input-output data is available. In this paper, a feed forward neural network that models the Fluid Catalytic Cracking (FCC) process will be presented. The FCC process is the workhorse of the petroleum refining industry, making small and medium sized molecules out of big ones (gasoline and distillate out of gas oils). The input-output data to the neural network was collected from the literature on pilot and commercial plant operations and were obtained from actual refineries. Several network architectures were tried and the network that best simulates the FCC process was retained. This network is able to predict yields of products of the FCC unit as well as their properties. The network consists of one hidden layer of twenty neurons, an input layer of four neurons, and an output layer of twelve neurons. The predictions of the neural network model were compared to those of a commercial simulator of the FCC process, to non-linear regression models, and to published charts. The results show that the neural network model consistently gives better predictions.     

An investigation on the sensitivity analysis of the parameters of proposed wax precipitation model

Reservoir hydrocarbon fluids contain heavy paraffins that may form solid phase of wax at low temperatures. Formation of solid phases is highly unwanted in oil production assemblies, pipelines and in process equipments. A predictive technique is crucial to the solution of wax formation to alleviate this problem.The effect of different parameters to predict the conditions under which wax precipitation takes place using the proposed model of Sahand University of Technology and other models has been investigated. The proposed model uses regular solution theory to describe solid phase (wax) non-ideality and the liquid and gas phases are being described by an equation of state.In order to evaluate the reliability of the proposed model, wax appearance temperatures (WAT's) were calculated for several mixtures at different compositions and compared with different models. The proposed model predictions had very good agreement with experimental data over a wide range of compositional distributions in comparison with other models. Solid wax content was also calculated at different temperatures below WAT in several synthetic systems made up of a solvent (decane) and a paraffinic heavy fraction. The results of calculating the amount of wax precipitation showed very good agreement with experimental data. Effect of different parameters including fusion temperature (T^f), Enthalpy of fusion (Δh^f), solubility parameter (δ^S), and binary interaction parameters (BIP) in predicting the WAT and the amount of wax precipitated for different oil mixtures have been evaluated using the proposed model and compared with other models. The results showed that the T^f is the most sensitive parameter while δ^S shows the least sensitivity in matching the WAT. Even though using Δh^f could provide the same results as tuning T^f, but the required changes are much higher and sometimes not practical. Also using BIP as the tuning parameter, requires a fairly large coefficient that makes it unsuitable to be considered as the tuning parameter.     

Estimation of wax deposition in the oil production units using RBF-ANN strategy

Accurate prediction of the solid deposition is petroleum industry can result in increasing the production efficiency. This can also result in the elimination of a major industrial problem, namely the wax deposition. In this study, application of intelligent methods in prediction of the wax deposition is investigated by developing a radial basis function artificial neural network. Levenberg Marquardt algorithm is also applied to determine the optimum predictions. Results from the proposed model are also compared to Kamari et al. model revealing the better performance of the proposed RBF-ANN. The validity of the proposed model is also investigated using statistical and graphical approaches, illustrating the great capability of the proposed RBF-ANN in accurate prediction of the wax deposition. R-squared and mean squared error values of 0.9975 and 0.029251 are obtained for the proposed model, revealing the validity of the RBF-ANN in estimating the wax deposition.     

The Prediction of Permeability Using an Artificial Neural Network System

Abstract

The authors studied the efficiency and accuracy of neural network model for prediction of permeability as a key parameter in reservoir characterization. So, some multilayer perceptron (MLP) neural network models with different learning algorithms of Levenberg-Margnardt, back propagation, improved back propagation (IBP), and quick propagation with three layers and different node numbers (3, 4, 5, 6, 7) in the middle layer have been presented. These models have been obtained by 630 permeability data from one of offshore reservoirs located in Saudi Arabia. The accuracy of models was studied by comparing the obtained results of each model with experimental data. So, the neural network with IBP learning method and five nodes in the middle layer has the most accuracy.     

Application of Radial Basis Function Neural Network for Prediction of Titration-Based Asphaltene Precipitation

The precipitation of asphaltene, a polar fraction of crude oil, during oil production has unfavorable impacts on many parts of the petroleum industry. Within the upstream processes, asphaltene precipitation occurs in crude oil, forming solid deposits in the reservoir during enhanced oil recovery operations and natural depletion. This significantly influences the porosity and permeability of the reservoir, reducing the effectiveness of the recovery process. Precipitation and deposition in downstream processes causes noticeable increases in production costs. Therefore, it is essential to predict the amount of asphaltene precipitation based on pressure, temperature and liquid phase composition using a dependable, precise, and robust strategy. However, the experimental measurement techniques used to estimate amounts are expensive and time consuming, while the thermodynamic models available are also somewhat complex. The authors propose an innovative approach for the simple and prompt prediction of asphaltene precipitation, employing an artificial neural network. The results show that the predicted values were in agreement with the experimental data, with the maximum absolute error deviation for the proposed model no more than 2.46%. A comparison of the proposed model with previously presented models highlight the superiority of the model developed in this study.     

An Artificial Neural Network Model for Design of Wellhead Chokes in Gas Condensate Production Fields

Abstract

The subcritical flow behavior of gas condensates through wellhead chokes under different flow conditions are studied by use of an artificial neural network (ANN). The proposed network is trained using the Levenberg-Marquardt back-propagation algorithm and the hyperbolic tangent sigmoid activation function is applied to calculate the output values of the neurons of the hidden layer. The proposed neuromorphic model outperforms the existing empirical correlations both in accuracy and generality. The results of this work are very important in the design of wellhead chokes under a wide range of flow conditions usually encountered during the flow of gas condensates.     

Artificial Neural Network's Prediction of Wax Deposition Potential of Nigerian Crude Oil for Pipeline Safety

Abstract

Paraffin wax deposition from crude oil along pipeline is a global problem, making preventive methods preferred to removal methods. In this work, a neural network model based on mathematical modeling technique using regression analysis as the statistical tool was developed to predict the wax deposition potential of 11 reservoirs in Nigeria. Using the viscosity-pressure-temperature data obtained from these fields to supervise the model, the model accurately predicted the present real-life situation in each field. Conclusively, the model could be used to predict wax deposition potential of any reservoir that is yet to be explored provided the temperature used during prediction is close to the actual reservoir temperature.     

Prediction of equilibrium water dew point of natural gas in TEG dehydration systems using Bayesian Feedforward Artificial Neural Network (FANN)

Abstract

The aim of this paper is to predict the equilibrium water dew point of natural gas in TEG dehydration process using feedforward artificial neural network (FANN). The ANN model shows a good result as the coefficient of determination of 0.9989 and 0.9976 was obtained for training and testing data respectively with relatively small value of mean square errors of 0.0203 and 0.0221. 0.5% of average absolute deviation percentage was observed which is comparable with the literatures. It clearly shows that FANN gives a good prediction on water dew point of natural gas in TEG dehydration process.     

Phase behavior modelling of asphaltene precipitation utilizing MLP-ANN approach

Since the sedimentation of heavy hydrocarbons such as asphaltenes, is the highlighted concern in production and operational, many studies were focused on this challenge in the petroleum industry. Therefore, the petroleum engineers should access to the asphaltene precipitation as an essential factor in order to conquer its problems. In this study, an empirical model for prediction asphaltene precipitation by multi-layer perceptron artificial neural network (MLP-ANN) is offered that takes the effect of the temperature, dilution ratio, and molecular weight for different n-alkanes. The output of this model showed 0.9999 for correlation coefficient (R²) and 0.000495 for mean squared error (MSE). This value illustrates the high quality of this model in compare of other available models. So far, MLP-ANN can offer significant accuracy in predicting asphaltene precipitation of asphaltene and other heavy oil.     

Predicting asphaltene precipitation during titration of diluted crude oil with paraffin using artificial neural network (ANN)

Abstract

Asphaltene precipitation from crude oil in underground reservoirs and on ground facilities is one of the major problems in a large portion of oil production units around the world. Many scaling equations and intelligent predictive models using the artificial neural network (ANN) are proposed in the literature but none of them can be applied when crude oil is diluted with any types of paraffin. In this study, feed forward artificial neural network is used for prediction of the amount of asphaltene precipitated weight percent of diluted crude oil with paraffin based on titration tests data from published literature. Trial and error method is utilized to optimize the artificial neural network topology in order to enhance its strength of generalization. The results showed that there is good agreement between experimental and predicted values. This predictive model can be applied to estimate the amount of asphaltene precipitated weight percent when the crude oil is diluted with paraffin and to avoid experimental measurement that is time-consuming and requires expensive experimental apparatus as well as complicated interpretation procedure.     

Development of an Artificial Neural Network Algorithm for the Prediction of Asphaltene Precipitation

Abstract

The precipitation and deposition of crude oil polar fractions such as asphaltenes in petroleum reservoirs considerably reduce rock permeability and oil recovery. Therefore, it is of great importance to determine how and how much the asphaltenes precipitate as a function of pressure, temperature, and liquid phase composition. The authors designed and applied an Artificial Neural Network (ANN) model to predict the amount of asphaltene precipitation at a given operating condition. Among this training, the back-propagation learning algorithm with different training methods was used. The most suitable algorithm with an appropriate number of neurons in the hidden layer, which provides the minimum error, was found to be the Levenberg-Marquardt (LM) algorithm. An extensive experimental data for the amount of asphaltene precipitation at various temperatures (293–343 K) was used to create the input and target data for generating the ANN model. The predicted results of asphaltene precipitation from the ANN model was also compared with the results of proposed scaling equations in the literature. The results revealed that scaling equations cannot predict the amount of asphaltene precipitation adequately. With an acceptable quantitative and qualitative agreement between experimental data and predicted amount of asphaltene precipitation for all ranges of dilution ratio, solvent molecular weight and temperature was obtained through using ANN model.     

Prediction of Asphaltene Precipitation in Live and Tank Crude Oil Using Gaussian Process Regression

The study of asphaltene precipitation properties has been motivated by their propensity to aggregate, flocculate, precipitate, and adsorb onto interfaces. The tendency of asphaltenes to precipitation has posed great challenges for the petroleum industry. Since the nature of asphaltene solubility is yet unknown and several unmodeled dynamics are hidden in the original systems, the existing models may fail in prediction the asphaltene precipitation in crude oil systems. The authors developed some Gaussian process regression models to predict asphaltene precipitation in crude oil systems based on different subsets of properties and components of crude oil. Using feature selection techniques they found some subsets of properties of crude oil that are more predictive of asphaltene precipitation. Then they developed prediction models based on selected feature sets. Results of this research indicate that the proposed predictive models can successfully predict and model asphaltene precipitation in tank and live crude oils with good accuracy.     

PREDICTING THE EFFECT OF FEEDSTOCK ON PRODUCT YIELDS AND PROPERTIES OF THE FCC PROCESS

The mechanism of petroleum refining processes are too complex, and no thorough model has yet been developed. Neural networks represent an effective alternative to mathematical modeling of refinery operations if a sufficient amount of input-output data is available. In this paper, a feed forward neural network that models the Fluid Catalytic Cracking (FCC) process will be presented. The FCC process is the workhorse of the petroleum refining industry, making small and medium sized molecules out of big ones (gasoline and distillate out of gas oils). The input-output data to the neural network was collected from the literature on pilot and commercial plant operations and were obtained from actual refineries. Several network architectures were tried and the network that best simulates the FCC process was retained. This network is able to predict yields of products of the FCC unit as well as their properties. The network consists of one hidden layer of twenty neurons, an input layer of four neurons, and an output layer of twelve neurons. The predictions of the neural network model were compared to those of a commercial simulator of the FCC process, to non-linear regression models, and to published charts. The results show that the neural network model consistently gives better predictions.     

Functional Link Neural Network for Predicting Crystallization Temperature of Ammonium Chloride in Air Cooler System

The air cooler is an important equipment in the petroleum refining industry. Ammonium chloride(NH_4 Cl) deposition-induced corrosion is one of its main failure forms. In this study, the ammonium salt crystallization temperature is chosen as the key decision variable of NH_4 Cl deposition-induced corrosion through in-depth mechanism research and experimental analysis. The functional link neural network(FLNN) is adopted as the basic algorithm for modeling because of its advantages in dealing with non-linear problems and its fast-computational ability. A hybrid FLNN attached to a small norm is built to improve the generalization performance of the model. Then, the trained model is used to predict the NH_4 Cl salt crystallization temperature in the air cooler of a sour water stripper plant. Experimental results show the proposed improved FLNN algorithm can achieve better generalization performance than the PLS, the back propagation neural network, and the conventional FLNN models.