Data Driven Prediction of the Minimum Miscibility Pressure (MMP) Between Mixtures of Oil and Gas Using Deep Learning

Abstract

Knowing the minimum miscibility pressure (MMP) between different oil and gas compositions is important to predict reservoir performance for gas-based injection as a secondary gas flood or tertiary technique such as water alternating gas (WAG). Machine Learning (ML) has been used widely and has been proven efficient in estimating these properties. In this work, the development of ML as well as commonly used algorithms in predicting bubble point pressure and oil formation volume factor is reviewed. Just a few studies are found before 2000. From 2001 to 2010, the use of ML increased steadily. However, a sharp augmentation in number of articles is observed from 2011 up to now. More than that, Artificial Neural Networks (ANN) is the most employed algorithm with 23 applications out of 38 studied papers. In addition, for the first time, deep learning- multiple fully connected networks algorithm is implemented to predict the MMP for oil and gas through 250 datasets covering a wide range of CO2 concentration from 0 to 100% in the injected gas. The wide range of CO2 concentrations is to cover all modes of gas injection from a pure CO2 flood to CO2 being negligibly present when injecting a sweet gas. The model is then optimized using Early Stopping and K-Fold Cross Validation techniques, showing the average result of k splitting data sets. The eight input parameters are as follows: reservoir temperature, oil characteristics (molecular weight, ratio of volatile components, and intermediate components), and gas characteristics (mole percentage of CO2, Cl, N2, H2S, C2+). The proposed model is compared with other Machine Learning Techniques such as Decision Tree and Random Forest Regression. The results show that reservoir temperature, the amount of CO2 and Cl in the gas source were the parameters to affect MMP the most significantly. The presence of CO2 in the gas stream will lower the MMP significantly. The Deep Learning model obtained an R2 = 0.96 and a Root Mean Square Error (RMSE) of 5.4%. Through Early Stopping technique, the proposed model reach the R2 result of 0.97 in 7 epochs. An R2 value of 0.954 was found using K-Fold Cross Validation technique, resulting in a good model generated by five folds data set. The model built by Deep Learning algorithm was more accurate than these ones built by Decision Tree and Random Forest Regression, which had an R2 value below 0.9 and RMSE larger than 10%. This work goes beyond other prior research by adding a ‘stopping point’ concept, increasing the overall performance of the methods for general applications, and considering the full range of CO2 in the gas stream.

Unveiling the Effect of Complex Components on CO2-Oil Minimum Miscibility Pressure: Insights from Deep Convolutional Neural Networks

“…Several other AI methods, including the BP NN with a single hidden layer, the SVM (support vector machine) algorithm with Gaussian and polynomial kernels, and the decision tree algorithm, have been previously employed for MMP value prediction. ,,, When they are applied to the testing data set, the prediction results are shown in Figure . For SVM using Gaussian and polynomial kernels, their predicted points closely align with the 45° line at lower MMP values but show obvious deviation at higher MMP values (Figure a,b).…”

“…22 In addition to the ML methods, neural network (NN) methods have also been applied increasingly into the MMP prediction. 9,23,24 For example, Chen et al 25 proposed an artificial neural network (ANN) model based on genetic algorithm, which considered molecular weight of pentane plus fraction, mole percentage of volatile and intermediate components in crude oil, and mole percentage of CO 2 , N 2 , CH 4 , H 2 S and intermediate hydrocarbon in gas. Sayyad et al 26 also developed an optimized ANN by particle swarm optimization, with the consideration of the same components.…”

“…However, due to the simple structure, most of these ML methods could only consider the limited number of components, e.g., pentane plus or heptane plus, volatile component, and intermediate components in crude oil. ,− Until recently, by improving the structure (using the conditional generative adversarial network together with the Bayesian optimization algorithm), the ML method could be capable of breaking the limit on the number of input attributes, predicting the MMP of the CO 2 -oil systems with more complex components . In addition to the ML methods, neural network (NN) methods have also been applied increasingly into the MMP prediction. ,, For example, Chen et al proposed an artificial neural network (ANN) model based on genetic algorithm, which considered molecular weight of pentane plus fraction, mole percentage of volatile and intermediate components in crude oil, and mole percentage of CO 2 , N 2 , CH 4 , H 2 S and intermediate hydrocarbon in gas. Sayyad et al also developed an optimized ANN by particle swarm optimization, with the consideration of the same components.…”

Utilizing Artificial Intelligence Techniques for Modeling Minimum Miscibility Pressure in Carbon Capture and Utilization Processes: A Comprehensive Review and Applications

“…The two models seemed to generate satisfactory performance, while the SVR based on the RBF kernel cannot reflect the impact of some input parameters on the investigated output. Pham et al 225 77 proposed five data-driven approaches for modeling the MMP of the systems CO 2 − oil using 155 data points. Their results revealed that ANN and SVR were the best models.…”

“…The two models seemed to generate satisfactory performance, while the SVR based on the RBF kernel cannot reflect the impact of some input parameters on the investigated output. Pham et al utilized DL to model the MMP for oil and gas systems, including pure and impure CO 2 (the concentration of CO 2 varies from 0 to 100% in their considered database). The gained model yielded interesting statistical metrics.…”