Introducing a New Method for Predicting PVT Properties of Iranian Crude Oils by Applying Artificial Neural Networks

Abstract

In this study artificial neural networks (ANNs) have been applied for the prediction of main pressure, volume, and temperature (PVT) properties, bubble point pressure (P_b), and bubble point oil formation volume factor (B_ob) of crude oil samples from different wells of Iranian oil reservoirs. Via a detailed comparison, the great power of ANNs with respect to traditional methods of predicting PVT properties, like Standing, Vasquez and Beggs, and Al-Marhoun, with higher prediction precision up to R² = 0.990 has been illustrated and the obtained parameters of ANNs for the application of prediction of other crude oil samples has been presented. The applied PVT data set in this study consists of 218 crude oil samples from Iranian reservoirs and for assurance of the applicability of the ANN model the PVT data set has been divided into 2 training (190 samples) and cross validation (28 samples) data sets and obtained ANNs from applying the training data set has been tested on the cross validation data set which has not been seen by the network during the training process. The obtained results for both training and cross validation data sets confirm the great prediction power of ANNs, for both data sets with respect to traditional PVT correlations.     

PVT-generated Correlations of Heavy Oil Properties

Pressure-volume-temperature (PVT) properties are necessary to reservoir engineering calculations in porous media and it is important for calculations in pipeline as well. This work presents a new set of correlation for estimating Iranian Crude oils properties based on some experimental data. Whenever there is no representative experimental PVT data, these correlations can be used for oils of API ranging between 15 and 30. New correlations was developed to calculate oil formation volume factor (Bo), bubble point pressure (Pb), and solution gas oil ratio (Rs). Finally, a comparison is made between these correlations and other published correlations such as Standing, Vazquez and Beggs, Glaso, Farshad et al., Al Marhoun, Petrosky and Farshad, and Hanafi et al. and it is found out that the new correlations are more accurate than the other ones.     

Predicting the Bubble-Point Pressure and Formation-Volume-Factor of Worldwide Crude Oil Systems

Abstract

This paper presents models for predicting the bubble-point pressure (P _b ) and oil formation-volume-factor at bubble-point (B _ob) for crude oil samples collected from several regions around the world. The regions include major producing oil fields in North and South America, North Sea, South East Asia, Middle East, and Africa. The model was developed using artificial neural networks with 5200 experimentally obtained PVT data sets. This represents the largest data set ever collected to be used in developing P _b and B _ob models. An additional 234 PVT data sets were used to investigate the effectiveness of the neural network models to predict outputs from inputs that were not used during the training process. The network model is able to predict the bubble-point pressure and the oil formation-volume-factor as a function of the solution gas–oil ratio, the gas relative density, the oil specific gravity, and the reservoir temperature. In order to obtain a generalized accurate model, back propagation with momentum for error minimization was used. The accuracy of the models developed in this study was compared in details with several published correlations. This study shows that if artificial neural networks are successfully trained, they can be excellent reliable predictive tools to estimate crude oil properties better than available correlations. The network models can be easily incorporated into any reservoir simulators and/or production optimization software.     

Comparison of the Performance of Empirical Models Used for the Prediction of the PVT Properties of Crude Oils of the Niger Delta

A review of the literature has revealed the lack of a formal analysis of the performance of empirical methods for the prediction of pressure volume temperature (PVT) properties of Niger delta crude oils. This study presents an assessment of the predictive accuracy of five bubble-point pressure (P_b) correlations and five bubble-point oil formation volume factor (B_ob) correlations against a large measured PVT data bank from Niger Delta crude oils. Statistical analysis techniques were used to evaluate the performance and the accuracy of the commonly used empirical models for estimating PVT properties of Niger crude oil in order to guide designers and operators in selecting the best correlations for their particular applications. Agreement between calculated and measured P_b and B_ob values for the various models was very poor. The model predictions of P_b and B_ob can be different from the measured values by 56% and 242%, respectively. Development of improved models for predicting the PVT properties of Niger delta crude oils is urgently required.     

A Correlations Approach for Prediction of PVT Properties of Reservoir Oils

Reservoir oil properties are usually measured at reservoir temperature and are estimated at other temperature using empirical correlations. Fluid properties correlations cannot be used globally because of different characteristics of fluids in each area. Here, based on Iranian oil PVT data, new correlations have been developed to predict saturation pressure and oil formation volume factor at bubble point pressure. Validity and accuracy of these correlations were confirmed by comparing results of these correlations with experimental data. Checking the results shows that results for Iranian oil properties in this work are in good agreement with experimental data respect to other correlations.     

Asphaltene Deposition under CO2 Injection: An Experimental Study of the Bangestan Reservoir of Ahwaz Oilfield,SW Iran

Abstract

It is essential that precipitation of asphaltenes is recognized early in the planning stage of any CO₂ enhanced oil recovery (EOR) project so that appropriate testing can be performed to evaluate whether there will be a negative impact on reservoir performance. This article presents detailed evaluations of slim tube data that were obtained during CO₂ injection using a medium-gravity Iranian crude oil.

A crude oil from Bangestan reservoir of Ahwaz oilfield containing 18.2% asphaltenes with ~31.5 °API gravity was flooded by purified CO₂ (>96% CO₂) in a slim tube apparatus under 2,700 psi at 110°C. We were going to determine the minimum miscibility pressure (MMP) of the sample oil under injection of CO₂ flood, but when a CO₂ slim tube test was performed for this oil at 2,700 psi, less than half of the saturated oil in the tube was recovered, which implied that the displacement process was immiscible. At this pressure, the asphaltene deposition in the slim tube apparatus was so severe that even a pressure gradient of 6,200 lb/in² was not able to displace any fluid through the capillary tube. Therefore, we abandoned MMP determination with this sample and investigated the problem.

Due to the high percentage of asphaltenes in the sample, using the slim tube MMP as an apparatus for determining minimum miscibility pressure of CO₂ and sample oil can be misleading.     

PVT EMPIRICAL MODELS FOR SATURATED OMANI CRUDE OILS

ABSTRACT

Pressure–volume–temperature (PVT) properties are critical to reservoir as well as production engineers, in particular. PVT properties could be determined experimentally. But, experiments are time consuming and costly. Moreover, laboratory PVT analysis does not consider the variations of PVT properties with respect to temperature since they are measured at reservoir temperature at the time of sampling. For that matter, the data is not benefitable. But, even if experimental analysis is done, it is difficult to obtain representative results to develop a new field. To tackle the above and other related problems, relying on sound PVT emperical correlations would be the ultimate solution. In this work, the intent is to develop stochastic models for PVT properties pertaining to Omani crude oils since it is believed that such correlations are scarce and not very precise. The empirical equations are developed for saturated Omani crude oils. The correlations are tested and validated. The empirical equations evaluation and assessment are done against existent experimental data and published correlations.     

ESTABLISHING PVT CORRELATIONS FOR OMANI OILS

ABSTRACT

An adequate knowledge of any reservoir fluid PVT properties is essential for most types of petroleum calculations. These calculations include amount of oil in the reservoir, production capacity, variations in produced gas-oil ratio during the reservoir's production life, calculation of recovery efficiency, reservoir performance, production operations and the design of production facilities. PVT properties can be measured experimentally by using collected bottom-hole or surface samples of crude oils. But, the experimental determination of PVT is time consuming and very costly. In addition, even with the availability of PVT analyses, it is often necessary to extrapolate the data to field and/or surface conditions through the use of empirical correlations. Furthermore, geological and geographical conditions are considered very critical in the development of any correlation. But, universal correlations are difficult to develop. That is why correlations for local regions, where crude properties are expected to be uniform, is a reasonable alternative. In this study, experimental PVT data for North and South Oman crudes, statistical and artificial neural network (ANN) analyses are used to develop reliable PVT correlations. Comparisons with previously published correlations are presented.     

Comparative evaluation of back-propagation neural network learning algorithms and empirical correlations for prediction of oil PVT properties in Iran oilfields

This paper presents a new approach to improve the performance of neural network method to PVT oil properties prediction. The true value of PVT properties which is determined based on the accurate data is a challenge of the petroleum industry. The main goal of the following investigation would be the performance comparison of various back-propagation learning algorithms in neural network that could be applied for PVT prediction. Up to now, no procedure has been presented to determine the network structure for some complicated cases, therefore; design and production of neural network would be almost dependent on the user's experience. To prevent this problem, neural network based recommended procedure in this study was applied to present the advantages. To show the performance of this procedure, several learning algorithms were investigated for comparison. One of the most common problems in neural network design is the topology and the parameter value accuracy that if those elements selection was correctly and optimally, the designer would achieve better results. Since, fluids of different regions have varying hydrocarbon properties, therefore, the empirical correlations in different hydrocarbon systems should be investigated to find their accuracies and limitations. In this study, an investigation of different empirical correlations along with the artificial neural networks in Iran oilfields has been presented. Then, the new model of artificial neural network for prediction of PVT oil properties in Iran crude oil presented. To test this new method, it was evaluated by collecting dataset from 23 different oilfields in Iran (south, central, western and continental shelf). In this study, two networks for prediction of bubble point pressure values (P_b) and the oil formation volume factor at bubble point (B_ob) were designed. The parameters and topology of the optimum neural networks were determined and in order to consider the effect of these networks designing on results, their performances were compared with various empirical correlations. According to comparison between the obtained results, it shows that the improved method presented has better performance rather than empirical and current methods in neural network designing in petroleum applications for these predictions.     

The Artificial Neural Network's Prediction of Crude Oil Viscosity for Pipeline Safety

Abstract

Predicting crude oil viscosity is a challenge faced by reservoir engineers in production planning. Some early researchers have propounded some theories based on crude oil properties and have encountered various problems leading to errors in forecasted values. This article discusses work carried out with a model using an artificial neural network (ANN) for predicting crude oil viscosity of Nigerian crude oil. The model was started through adoption of a classical regression technique empirical method for dead oil viscosity as a function of American Institute for Petroleum (API) and reduced temperature. The Peng–Robinson equation of state and other thermodynamic properties are introduced, coupled with the Standing model for calculating bubble point pressure (P_b). The developed model was evaluated using existing measured real-life data collected from 10 oil fields within the Niger Delta region of Nigeria. Both the predicted and measured viscosities were plotted against each corresponding reservoir pressure to establish the model's level of reliability. The superimposition of the pressure-viscosity relationship shows that at each point, the viscosity model captures the physical behavior of viscosity variations with pressure. In each case, the ANN does not require a data relationship to predict the crude oil viscosity but rather relies on the field data obtained for training. For this reason, it is recommended that the ANN approach should be applied in oil fields for reduction in error, computational time, and cost of overproduction and underproduction.     

Predicting Bubble-point Pressure and Formation-volume Factor of Nigerian Crude Oil System for Environmental Sustainability

Abstract

This paper presents a model for predicting the bubble–point pressure (P _b ) and oil formation-volume-factor at bubble-point (B _ob ) for crude samples collected from some producing wells in the Niger-Delta region of Nigeria. The model was developed using artificial neural networks with 542 experimentally obtained Pressure-Volume-Temperature (PVT) data sets. The model accurately predicts the P _b and B _ob as functions of the solution gas-oil ratio, the gas relative density, the oil specific gravity, and the reservoir temperature. In order to obtain a generalized accurate model, backpropagation with momentum for error minimization was used. The accuracy of the developed model in this study was compared with some published correlations. Apart from its accuracy, this model takes a shorter time to predict the PVT properties when compared with empirical correlations. The immediate reason for this may have to do with the non-linear nature of the empirical correlations.     

Influence of the Microwave Method on Vanadium Removal in Crude Oil

Abstract

Microwave technology was introduced for removing vanadium from crude oil. This article discusses the effect of the microwave on removing vanadium from Iranian and Shengli crude oil. We investigated the impact of the microwave time, initial temperature, and microwave power on removing vanadium from crude oil. The optimal conditions of vanadium removal from Iranian and Shengli crude oil were as follows: (a) microwave power was 350 W, (b) microwave time was 30 min, and (c) initial temperature had little influence on the removal efficiencies of vanadium. Under these conditions, the removal rates of vanadium from Iranian and Shengli crude oil were 92.00% and 88.00%, respectively.     

CFD investigation on characteristics of heavy crude oil flow through a horizontal pipe

Transportation of heavy crude oil via pipelines possesses many technological issues that are inherently flow related. Accurate prediction of flow characteristics is an essential step for a reliable piping design of transporting the crude oil. A rheology-based Computational Fluid Dynamics (CFD) model of the Iraqi heavy crude oil flow through a horizontal pipe (1 m length of 3/4 in. inside diameter) was developed using the commercial software Ansys 15 Fluent. By using power law rheological model, the Iraqi heavy crude oil exhibits a non-Newtonian dilatant behavior over the examined shear rate range of 1–40 s⁻¹. The proposed axi-symmetric CFD model identifies velocity profile and generates values of friction factor, which are validated with experimental measurements. Additionally, wall shear stress and entrance length were numerically predicted and compared with well-established correlations from the literature for Non-Newtonian flow. Detailed results of the CFD model exhibited a reliable prediction of the characteristics of heavy crude oil flow.     

A Rigorous Yet Simple Dead Oil Viscosity Correlation

A rigorous yet simple correlation for the estimation of dead oil viscosity is proposed. The new correlation requires oil API gravity and system temperature as the only correlation parameters. It calculates the Watson characterization factor as a function of oil API gravity. Hence, the paraffinicity or character of the crude oil is implicitly taken into account. The new correlation was checked against other correlations for a full spectrum of oil API gravities and system temperatures. It performed exceptionally good thus eliminating the lengthy and complex calculation procedure of other correlations.     

Analyzing and assessment of oil viscosity models based on the API ranges

The common empirical correlations of oil viscosity can be divided into three categories: extra-heavy oil models, heavy oil models and light oil models. But not all the models are suitable for predicting these three kinds of oil. Each crude oil viscosity model has the best API scope of use. Therefore, this paper collected 293 oil samples from the published articles, and these samples were divided into 4 categories according to the API (API ≤ 10, 10 < API ≤ 20, 20 < API ≤ 30, API > 30). In this paper, the accuracy and reliability of each model at different API ranges are assessed and evaluated. In addition to the modification of the commonly used models, the new models are also recommended at different API scopes. The new prediction models are precise than other common models, calculated values and measured values coincide when degree is higher. It is very important to figure out which model is best in different API ranges.     

Viscosity prediction in selected Iranian light oil reservoirs: Artificial neural network versus empirical correlations

Viscosity is a parameter that plays a pivotal role in reservoir fluid estimations. Several approaches have been presented in the literature (Beal, 1946; Khan et al, 1987; Beggs and Robinson, 1975; Kartoatmodjo and Schmidt, 1994; Vasquez and Beggs, 1980; Chew and Connally, 1959; Elsharkawy and Alikhan, 1999; Labedi, 1992) for predicting the viscosity of crude oil. However, the results obtained by these methods have significant errors when compared with the experimental data. In this study a robust artificial neural network (ANN) code was developed in the MATLAB software environment to predict the viscosity of Iranian crude oils. The results obtained by the ANN and the three well-established semi-empirical equations (Khan et al, 1987; Elsharkawy and Alikhan, 1999; Labedi, 1992) were compared with the experimental data. The prediction procedure was carried out at three different regimes: at, above and below the bubble-point pressure using the PVT data of 57 samples collected from central, southern and offshore oil fields of Iran. It is confirmed that in comparison with the models previously published in literature, the ANN model has a better accuracy and performance in predicting the viscosity of Iranian crudes.     

Enhanced oil recovery by means of alkali injection. Behavior of the SARA fractions

Abstract

The presence of an alkali such as sodium hydroxide (NaOH) within a process of enhanced recovery of crude oil having certain acidity allows the formation of surfactants with the ability of enhancing the separation of the crude oil from the reservoir rock where it is adsorbed due to a change of wettability. This work focused on studying the effect of NaOH on the enhanced recovery, in terms of the SARA composition, of two different API gravity crude oils (38.5 and 24.1°) having different total acid number. The results showed the existence of acid components with high and low reactivity to NaOH in the crude oil, where the first group is responsible of the in-situ synthesis of the surfactant. Moreover, it was found that the utilization of the alkali improved the recovery of the SARA fractions for each crude oil, especially for the 24.1°API crude oil for which the recovery resulted twice or three times higher to the evaluated for the 38.5°API crude oil. Additionally, it was found that a minimum of 0.2?wt% of NaOH reduced up to 4 times the value of the interfacial tension between the crude oil and the brine, regardless of the type of crude oil.     

Flow Enhancement of Medium-Viscosity Crude Oil

Abstract

This study investigated the different alternatives to enhance the flowability of crude oil with medium viscosity. These alternatives include the addition of water into crude oil to form water-in-oil emulsion, the addition of light petroleum product, the addition of flow improver, and a preheating technique. Temperature range of 10–50°C, water concentration range of 0–50% by volume, flow improver concentration range of 0–5000 ppm, and kerosene concentration range of 0–50% by volume were investigated in the flowability enhancement study of crude oil with medium viscosity. The flowability enhancement in terms of viscosity reduction was investigated using RheoStress RS100 from Haake. A cone–plate sensor was used with a cone angle of °4, cone diameter of 35 mm, and 0.137-mm gap at the cone tip. The addition of kerosene to crude oil improves the flowability much better than any other investigated technique.     

Improved Correlations for Fluid Properties of UAE Crude Oils

Abstract

Empirical correlations to evaluate crude oil fluid properties such as the formation volume factor, bubble point, viscosity, and oil compressibility above the bubble point are used extensively by petroleum and process engineers to perform calculations for subsurface and surface processes. The published correlations are mostly based on regional data, such as Standing's for California crudes, Petrosky and Farshad's for Gulf of Mexico crudes, and Glaso's for North Sea crudes. Use of these regional correlations is more appropriate for crudes from the same basins for which the correlation is derived. Other correlations, such as the Vasquez and Beggs correlation, are based on data from a very large number of samples coming from multiple regions. Eventhough one is tempted to use these “universal” correlations, the range of error for their predictions is, however, typically large due to the scatter involved in using a large number of data sets to generate these correlations. The UAE fields are quite significant and constitute around 9% of worldwide reserves. In this work, experimental PVT measurements from 15 medium to large fields located in the UAE are used to test the viability of using either regional or universal correlations to UAE crudes. In addition, a new set of empirical correlations is constructed based on these data, and their predictions are also compared. Statistical comparisons indicate that the new correlations developed in this paper reduce the error involved in predicting the bubble point pressures, the oil compressibility, and the oil formation volume factor to less than half the range associated with either the regional or the universal correlations. Also, new correlations for viscosity both at the bubble point and above the bubble point were constructed that gave better predictions for UAE oils over other commonly used correlations for viscosity.