Search

Shear wave velocity in the company of compressional wave velocity add up to an invaluable source of information for geomechanical and geophysical studies. Although compressional wave velocity measurements exist in almost all wells, shear wave velocity is not recorded for most of elderly wells due to lack of technologic tools in those days and incapability of recent tools in cased holes. Furthermore, measurement of shear wave velocity is to some extent costly. This study proposes a novel methodology to remove aforementioned problems by use of support vector regression tool originally invented by Vapnik (1995, The Nature of Statistical Learning Theory. Springer, New York, NY). Support vector regression (SVR) is a supervised learning algorithm plant based on statistical learning (SLT) theory. It is used in this study to formulate conventional well log data into shear wave velocity in a quick, cheap, and accurate manner. SVR is preferred for model construction because it utilizes structural risk minimization (SRM) principle which is superior to empirical risk minimization (ERM) theory, used in traditional learning algorithms such as neural networks. A group of 2879 data points was used for model construction and 1176 data points were employed for assessment of SVR model. A comparison between measured and SVR predicted data showed SVR was capable of accurately extract shear wave velocity, hidden in conventional well log data. Finally, a comparison among SVR, neural network, and four well-known empirical correlations demonstrated SVR model outperformed other methods. This strategy was successfully applied in one of carbonate reservoir rocks of Iran Gas-Fields.

Saturation pressure is one of the foremost parameters of crude oil which plays a key role in petroleum calculations. Experimentally, determination of this parameter in laboratory is costly and labor demanding. In this study, an improved intelligent model based on neural network optimized with genetic algorithm-pattern search technique is proposed for building quantitative formulation between saturation pressure and compositional data, including temperature, hydrocarbon and non-hydrocarbon compositions of crudes, and heptane-plus specifications. Genetic algorithm-pattern search technique is embedded in neural network formulation for finding optimal weights and biases of neural network. A comparison among the proposed model and published models in literature reveals the superiority of our model in terms of better accuracy and higher generalization. Improved neural networked showed R-square of 0.9892 and MSE of 17,617.99 which concludes that it is a promising alternative for determination of saturation pressure which is able to eliminating expenses of laboratory measurements and significantly saving time. This study showed GA considerably enhanced performance of conventional neural networks.

Bubble point pressure (Pb), one of the most important parameters of reservoir fluids, plays an important role in petroleum engineering calculations. Accurate determination of Pb from laboratory experiments is time, cost and labor intensive. Therefore, the quest for an accurate, fast and cheap method of determining Pb is inevitable. In this communication, a sophisticated approach was followed for formulating Pb to temperature, hydrocarbon and non-hydrocarbon compositions of crudes, and heptane-plus specifications. Firstly, support vector regression (SVR), a supervised learning algorithm plant based on statistical learning (SLT) theory, was employed to construct a model estimating Pb. Subsequently, an alternating conditional expectation (ACE) was used to transform input/output data space to a highly correlated data space and consequently to develop a strong formulation among them. Eventually, SVR and ACE models are combined in a power-law committee machine structure by virtue of genetic algorithm to enhance accuracy of final prediction. A comparison among constructed models and previous models using the concepts of correlation coefficient, mean square error, average relative error and absolute average relative error reveals power-law committee machine outperforms all SVR, ACE, and previous models.

Shear wave velocity provides invaluable information for geomechanical, geophysical, and reservoir characterization studies. However, measurement of shear wave velocity is time, cost and labor intensive. This study proposes a swift and exact methodology, called ACE stimulated neural network, for prediction of shear wave velocity from available well logs such that it will be able to surpass previous models. The proposed method is composed of two major parts: 1) transforming input/output data space to a higher correlated space using alternative condition expectation (ACE), and 2) making a neural network formulation in transformed data space. Transforming in the first step makes it easier for neural network to find the complicated underlying dependency of input/output data. Therefore, neural network will be able to develop an accurate and strong formulation between conventional well logs and shear wave velocity. The Propounded approach was successfully applied in one of the carbonate gas fields of Iran. A comparison between proposed model and previous models showed superiority of ACE stimulated neural network.

Poisson's ratio is one of the most important rock mechanical parameters having significance in both planning and post analysis of wellbore operations. Laboratory measurement of this parameter covers a broad range of costs, including sidewall sampling, preservation, and laboratory tests. This study proposes an improved strategy, called dual stimulated fuzzy logic by ACE and GA-PS for determining Poisson's ratio from conventional well log data in a rapid, precise, and cost-effective way. Firstly, conventional well log data are transformed to a higher correlated data space with Poisson's ratio through the use of alternative condition expectation (ACE) algorithm. This step simplifies the convoluted space of the problem and makes it easier to solve for fuzzy logic. Subsequently, transformed conventional well log data are fed to fuzzy logic model. To ensure that optimal fuzzy model is constructed, a hybrid genetic algorithm-pattern search (GA-PS) technique is employed for extracting fuzzy clusters (or rules). This step sets fuzzy logic to its optimal performance. The propounded strategy was successfully applied to data from carbonate reservoir rocks of an Iranian Oil Field. A comparison between present model and previous models showed superiority of current study.

Quantitative formulation between conventional well log data and rock permeability, undoubtedly the most critical parameter of hydrocarbon reservoir, could be a potent tool for solving problems associated with almost all tasks involved in petroleum engineering. The present study proposes a novel approach in charge of the quest for high-accuracy method of permeability prediction. At the first stage, overlapping of conventional well log data (inputs) was eliminated by means of principal component analysis (PCA). Subsequently, rock permeability was predicted from extracted PCs using multi-layer perceptron (MLP), radial basis function (RBF), and generalized regression neural network (GRNN). Eventually, a committee neural network (CNN) was constructed by virtue of genetic algorithm (GA) to enhance the precision of ultimate permeability prediction. The values of rock permeability, derived from the MPL, RBF, and GRNN models, were used as inputs of CNN. The proposed CNN combines results of different ANNs to reap beneficial advantages of all models and consequently producing more accurate estimations. The GA, embedded in the structure of the CNN assigns a weight factor to each ANN which shows relative involvement of each ANN in overall prediction of rock permeability from PCs of conventional well logs. The proposed methodology was applied in Kangan and Dalan Formations, which are the major carbonate reservoir rocks of South Pars Gas Field-Iran. A group of 350 data points was used to establish the CNN model, and a group of 245 data points was employed to assess the reliability of constructed CNN model. Results showed that the CNN method performed better than individual intelligent systems performing alone.

The portion of rock pore volume occupied with non-hydrocarbon fluids is called water saturation, which plays a significant role in reservoir description and management. Accurate water saturation, directly measured from special core analysis is highly expensive and time consuming. Furthermore, indirect measurements of water saturation from well log interpretation such as empirical correlations or statistical methods do not provide satisfying results. Recent works showed that fuzzy logic is a robust tool for handling geosciences problems which provide more reliable results compared with empirical correlations or statistical methods. This study goes further to improve fuzzy logic for enhancing accuracy of final prediction. It employs hybrid genetic algorithm-pattern search technique instead of widely held subtractive clustering approach for setting up fuzzy rules and for extracting optimal parameters involved in computational structure of fuzzy model. The proposed strategy, called genetic implanted fuzzy model, was used to formulate conventional well log data, including sonic transit time, neutron porosity, formation bulk density, true resistivity, and gamma ray into water saturation, obtained from subtractive clustering approach. Results indicated genetic implanted fuzzy model performed more satisfyingly compared with traditional fuzzy logic model. The propounded model was successfully applied to one of Iranian carbonate reservoir rocks.

Deposition of asphaltene during crude oil production is a challenging issue in oil industry which causes considerable loss of production efficiency as well as imposes negative impacts on production rates. Upon variation in pressure, temperature and crude oil composition, asphaltene begins to precipitate and deposits in reservoir rock and consequently causes formation damage owing to mechanisms of wettability alteration and pore throat blockage. In the present study a sophisticated method, called committee machine with optimized intelligent systems was utilized to predict the amount of asphaltene precipitation from experimental titration data. The committee machine is composed of optimized neural network and optimized fuzzy logic. Stochastic optimization of neural network and fuzzy logic by virtue of hybrid genetic algorithm-pattern search technique significantly enhances their efficiencies. The committee machine provides a further improvement in accuracy of final prediction through integrating optimized intelligent systems and consequent reaping of their benefits. The committee machine model was applied to experimental data reported in the open-source literature. It was observed that there was an acceptable agreement between experimental data and committee machine predicted values. Finally, performance of committee machine model was compared with other intelligent systems used for prediction of asphaltene precipitation. Results showed superiority of committee machine in asphaltene precipitation modeling to optimized neural network and optimized fuzzy logic.

Porosity, the void portion of reservoir rocks, determines the volume of hydrocarbon accumulation and has a great control on assessment and development of hydrocarbon reservoirs. Accurate determination of porosity from core analysis is highly cost, time, and labor intensive. Therefore, the mission of finding an accurate, fast and cheap way of determining porosity is unavoidable. On the other hand, conventional well log data, available in almost all wells contain invaluable implicit information about the porosity. Therefore, an intelligent system can explicate this information. Fuzzy logic is a powerful tool for handling geosciences problem which is associated with uncertainty. However, determination of the best fuzzy formulation is still an issue. This study purposes an improved strategy, called hybrid genetic algorithm–pattern search (GA–PS) technique, against the widely held subtractive clustering (SC) method for setting up fuzzy rules between core porosity and petrophysical logs. Hybrid GA–PS technique is capable of extracting optimal parameters for fuzzy clusters (membership functions) which consequently results in the best fuzzy formulation. Results indicate that GA–PS technique manipulates both mean and variance of Gaussian membership functions contrary to SC that only has a control on mean of Gaussian membership functions. A comparison between hybrid GA–PS technique and SC method confirmed the superiority of GA–PS technique in setting up fuzzy rules. The proposed strategy was successfully applied to one of the Iranian carbonate reservoir rocks.

Poisson ratio is considered as one of the most important rock mechanical properties of hydrocarbon reservoirs. Determination of this parameter through laboratory measurement is time, cost, and labor intensive. Furthermore, laboratory measurements do not provide continuous data along the reservoir intervals. Hence, a fast, accurate, and inexpensive way of determining Poisson ratio which produces continuous data over the whole reservoir interval is desirable. For this purpose, support vector regression (SVR) method based on statistical learning theory (SLT) was employed as a supervised learning algorithm to estimate Poisson ratio from conventional well log data. SVR is capable of accurately extracting the implicit knowledge contained in conventional well logs and converting the gained knowledge into Poisson ratio data. Structural risk minimization (SRM) principle which is embedded in the SVR structure in addition to empirical risk minimization (EMR) principle provides a robust model for finding quantitative formulation between conventional well log data and Poisson ratio. Although satisfying results were obtained from an individual SVR model, it had flaws of overestimation in low Poisson ratios and underestimation in high Poisson ratios. These errors were eliminated through implementation of fuzzy classifier based SVR (FCBSVR). The FCBSVR significantly improved accuracy of the final prediction. This strategy was successfully applied to data from carbonate reservoir rocks of an Iranian Oil Field. Results indicated that SVR predicted Poisson ratio values are in good agreement with measured values.

Bubble point pressure is the most crucial pressure–volume–temperature (PVT) property of reservoir fluid, which plays a critical role in almost all tasks related to reservoir and production engineering. Therefore, an accurate, quick, and easy way of predicting bubble point pressure from available PVT parameters is desired. In this study, an improved methodology is followed for making a quantitative formulation between bubble point pressure (target) and some available PVT data (inputs) such as proportion of solution gas–oil-ratio over gas gravity, temperature, and stock-tank oil gravity. At the first stage of this research, bubble point pressure was predicted from PVT data using different intelligent systems, including neural network, fuzzy logic, and neuro-fuzzy algorithms. Subsequently, a power-law committee with intelligent systems was constructed by virtue of hybrid genetic algorithm-pattern search tool. The proposed methodology, power-law committee with intelligent systems, comprises a parallel framework that produces a final output by combining the results of individual intelligent systems. To achieve this objective, a power-law formula structure was designated to integrate outputs of intelligent systems. A hybrid genetic algorithm-pattern search tool was then employed to find the optimal coefficients of this formula. A database of 361 worldwide data points was employed in this study, while 282 data points were used for model construction (i.e., training data), and 79 data points were employed to assess the reliability of the model (test data). Results showed that outputs of intelligent systems are in good agreement with reality. However, by little additional computation, power-law committee with intelligent systems is capable of significantly improving the accuracy of target prediction.