Ali Elkamel

THE ACCURACY OF PREDICTING COMPRESSIBILITY FACTOR FOR SOUR NATURAL GASES

This paper presents the initial stage of an effort aimed at developing a new correlation to estimate pseudo critical properties for sour gas when the exact composition is not known. Several mixing rules and gas gravity correlations available in the literature are first evaluated and compared. The evaluation is performed on a large database consisting of more than 2106 samples of sour gas compositions collected worldwide. Several evaluation criteria are used including the average absolute deviation (AAD), the standard deviation (SD), the coefficient of correlation, R, and cross plots and error histograms. The mixing rules include: Kays mixing rule combined with Wichert–Aziz correlation for the presence of non-hydrocarbons, SSBV mixing rule with Wichert and Aziz, Corredor et al. mixing rule, and Piper et al. mixing rule. These methods, in one form or another, use information on gas composition. Three different other methods that are based on gas gravity alone were also analyzed. These are: Standing, Sutton, and Elsharkawy et al. gas gravity correlations. While the methods based on knowledge of composition showed reasonable accuracy, those based on gas gravity alone showed weak accuracy with low correlation coefficients. A new gas gravity correlation that is based on the fraction of non-hydrocarbons present in the sour gas was proposed. Preliminary results indicate that a good improvement over past gravity correlations was achieved. The compositional correlations, still show, however, better accuracy. Research is still going on to come up with more accurate correlations that are based on only readily available descriptors.

An artificial neural network for predicting and optimizing immiscible flood performance in heterogeneous reservoirs

Enhanced Oil Recovery (EOR) processes are used to recover additional oil left in place after primary recovery and water flooding stages. The prediction of their performance is of great importance in the selection and design of a certain EOR process and also during planning for oil production. This work presents an extension of an earlier attempt on the use of neural networks to predict reservoir performance in homogeneous reservoirs. The consideration of heterogeneity and its accompanying interactions with the fluid flow equations in a porous media is considered in this paper. The independent dimensionless groups that characterize the flow behavior in a heterogeneous media have been used as inputs to a neural network model in order to predict oil recoveries. Various neural network architectures have been considered, and the network that best mimics a reservoir numerical simulator was retained. The simulations of this network are compared to those obtained from the reservoir simulator. The effect of various dimensionless groups on the oil recovery is discussed and an optimization study is performed to determine from the prepared neural network the optimal conditions leading to the best oil recovery efficiencies.

MODELING THE HYDROCRACKING PROCESS USING ARTIFICIAL NEURAL NETWORKS

ABSTRACT Feed-forward neural networks that models the hydrocracking process of Arabian light vacuum gas oil are presented. The input-output data to the neural networks was obtained from actual local refineries. Several network architectures were tried and the networks that best simulate the hydrocracking process were retained. The networks are able to predict yields and properties of products of the hydrocracking unit (e.g. iC4, nC4, light and heavy naphtha, light and heavy ATK, Diesel, etc.). The predictions of yields and properties of various desired and undesired products at different conditions are required by refineries for process optimization, control, design, catalyst selection, and planning. The predictions of the prepared neural networks have been cross validated against data not originally used in the training process. The networks compared well against this new set of data with an average percent error always less than 8.71 for the different products of the hydrocracking unit.

Producing ultralow interfacial tension at the oil/water interface

ABSTRACT In view of the world-wide shortage of petroleum and the fact that a large amount of residual oil will remain in the reservoir after the primary recovery and water flooding stages, the use of Enhanced Oil Recovery (EOR) methods to recover as much as possible of this residual oil has become increasingly important worldwide. The predominant and most promising EOR technique is the micellar-polymer flooding process which uses a surface active agent (a surfactant) to decrease interfacial tension and hence allows oil to freely move from its original location through the porous media. The purpose of this paper is to present an experimental study of the factors affecting the equilibrium interfacial tension (IFT) at the oil/water interface. A large number of experiments was conducted to study the variations of IFT as a function of many parameters including reservoir temperature, pressure, surfactant concentration, and salinity. An Arabian heavy crude oil was used in the analysis along with three different synthetic surfactants and two formation waters. The pendent drop technique enhanced by video imaging was employed for measuring IFT. It was found that for the ranges of variables considered in this study, IFT decreases with temperature and salinity, increases with pressure, and decreases exponentially with surfactant concentration.

Modelling pressure distribution in a rectangular gas bearing using neural networks

Abstract Gas ubricated bearings are of tremendous use especially in the biomedical and aerospace industries. For that reason, gas bearings have been the subject of much research for the past decade or so. Experimental as well as theoretical work has been done to calculate the pressure distribution inside the bearing. The models available to predict the pressure are primitive and need to be improved. This paper discusses a new modelling scheme known as artificial neural networks. The pressure distribution and the load-carrying capacity are predicted using feedforward architecture of neurons. The inputs to the networks are a collection of experimental data. This data is used to train the network using the Levenberg-Marquardt optimization technique. The results of the neural network model are compared to a theoretical model and the results are promising. The neural network model outperforms the avallable theoretical model in predicting the pressure as well as the load-carrying capacity.

Homopolymerization of 4‐propionoxybenzoic acid: A kinetic study

A kinetic study of the synthesis of poly(4-oxybenzoate) by melt-step growth polymerization using para-propionoxybenzoic acid is reported. The polycondensations obey second-order kinetics, irrespective of whether the reaction was catalyzed or uncatalyzed. Breaks are observed in the kinetic plots, suggesting the presence of different kinetic regimes during the course of the reaction. An elaborate kinetic model that presupposes precipitation of oligomers predicts two-stage kinetics as well as breaks in the rate plots and fits experimental data well throughout the course of the reaction and the performance of two transesterification catalysts are estimated. No isokinetic temperature is displayed for the transesterification reaction. Activation energy values for catalyzed reactions are found to be higher than the uncatalyzed reaction, indicating that entropy factors drive the reaction to completion.

STUDYING THE INTERACTIONS BETWEEN AN ARABIAN HEAVY CRUDE OIL AND ALKALINE SOLUTIONS

ABSTRACT An experimental study to examine the effectiveness of alkaline flooding for the recovery of an Arabian heavy crude oil is presented. The interfacial tension (IFT) behavior of crude oil/alkali systems over a wide range of parameters (pressure, temperature, alkali concentration and time) was studied. These alkaline reagents react with the acidic species in crude oil to form surface-active soaps in-situ. This leads to a lowering of interfacial tension (IFT) and subsequently the mobilization of residual oil. The equilibrium IFTs obtained through alkaline flooding are compared with the IFTs when a synthetic surfactant (dodecyl benzene sulfonic acid sodium salt) is used in EOR recovery. A mathematical model representing the complete chemistry of the transient process is also presented. The model consists of a set of differential equations describing reactions, diffusion, and adsorption at the oil/alkaline solution interface. The kinetic parameters in the model are obtained through a differential algebraic optimization technique. The concentration of the surface active species are related to the measured IFTs through an independent step that is based on isolating the surface active species formed by the reaction between the acids in the crude oil and the alkaline solution. A sensitivity analysis using the model is carried out to study the effect of surface potential and alkaline concentration on the transient interfacial tensions.

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.

DEVELOPMENT OF REGRESSION MODELS TO CONTROL PRODUCT YIELDS AND PROPERTIES OF THE FLUID CATALYTIC CRACKING PROCESS

Regression models for predicting product yields and properties for the Fluid Catalytic Cracking (FCC) process are presented. The feedstock properties are used as the primary correlation parameters. The models are based on data collected from various sources from the literature on pilot and commercial plants. The predictions of these models are compared to previous models and also to real data. The proposed models give consistently good predictions. These models can be used to evaluate different feeds to the FCC process and can also be integrated within general refinery mathematical programming (e.g. LP) software for planning and scheduling purposes.

A neural network prediction model of fluid displacements in porous media

This paper presents the development and design of an artificial neural network that is able to predict the breakthrough oil recovery of immiscible displacement of oil by water in a two-dimensional vertical cross section. The data used in training the neural network was obtained from the results of fine-mesh numerical simulations. Several network architectures were investigated and trained using the back propagation with momentum algorithm. The neural network that gave the best predictive performance was a two-hidden layer network with 8 neurons in the first hidden layer and 8 neurons in the second hidden layer. This network also performed well against a cross validation test. The reservoir simulation data used so far in the training process was for a homogeneous reservoir, the more general case is still under investigation.