Integrated Machine Learning Model for Predicting Asphaltene Damage Risk and the Asphaltene Onset Pressure

Abstract

Asphaltene precipitation can promote a drastic reduction in oil production because of asphaltene precipitation and deposition damage. Therefore, screening models to predict the risk of asphaltene damage and equations of state (EoS) to predict the asphaltene onset pressure (AOP) are useful to prevent production drops and optimize the management of oil resources. Most asphaltene screening models have been focused on the oil compositions (SARA analysis); however, these screening models do not consider key variables for asphaltene stability such as the temperature, pressure, well depth, gas–oil ratio, and so on. As the EoS are typically based on experimental data, to fit parameters needed to reproduce the experimental AOP, expensive laboratory analyses are required for this objective. In this study, a classification machine learning (CML) model based on support vector machines was proposed to predict the asphaltene damage risk from the asphaltene stability class index data and the in situ live crude oil densities. In addition, a model based on linear regression (MLR) to predict AOP from the reservoir pressure, saturation pressure, temperature, and some in situ live crude oil compositions was proposed. In total, 24 crude oils were evaluated experimentally to propose the classification model, and a perfect classification accuracy (100%) was obtained in both cases. The CML results were compared to compositional screenings, where classification accuracy was between 29 and 88%, the best accuracy being obtained from the well-known de Boer plot. The MLR model was obtained from data from 53 live crude oils, using hypothesis tests to select the statistically representative characteristics regarding AOP, and a determination coefficient of 0.77 was obtained. The proposed integrated ensemble model contributes to predicting the potential risk of damage due to asphaltene precipitation and estimating a pressure range where these asphaltenes precipitate, allowing the necessary preventive measures to be taken to avoid an oil production decline.

Structural-Compositional Effect of Polycyclic Aromatic Hydrocarbons on Thermal-Oxidative Decomposition at High-Pressure: A Molecular Dynamics and Machine Learning Approach

“…The variables that were found to be strongly correlated with the output variable and are statistically significant ( p -value test) and linearly independent between them were used to train the ML model. Typically, 80% of the data set is used for model training, with the remaining 20% used for validation; ,, this corresponds to 24 training data points and six validation data points. However, considering that each mass conversion profile is composed of five data points, we decided to use 25 data points to train the ML model and five data points for the validation of the model, i.e., we used 83–17% training–validation split.…”

Expansion of Nanotechnology-Based Chemically Enhanced Oil Recovery Field Applications in Colombia: An Approach from Laboratory Experiments, Effluent Follow-Up, and Machine Learning

“…From all of the data, 80% was used to train the model, and the remaining 20% served as the test subset to evaluate its performance on data not seen during training, thus assessing its predictive capacity. The choice of the 80–20 ratio for the division into training and test sets is a convention commonly used in modeling from ML. ,− …”

“…Data science and ML tools have been very useful in the oil and gas industry to develop reliable models that describe and forecast variables of interest in complex systems. Thus, the specialized literature on the subject has shown how the application of data science and machine learning techniques, such as artificial neural networks and random forests, has improved the accuracy and performance in predicting variables related to oil production. − These approaches offer new opportunities to optimize extraction processes and significantly impact the oil and gas industry. − However, it is worth mentioning that machine learning techniques have not been used so far to predict oil production data from oil fields based on the analysis of actual production and effluents due to limited applications of NanoCEOR under real oilfield conditions.…”

Machine Learning in Complex Organic Mixtures: Applying Domain Knowledge Allows for Meaningful Performance with Small Data Sets

“…Recent ML applications toward asphaltenes have typically attempted to predict bulk properties (e.g., precipitation/stability) from bulk features (e.g., “SARA” saturates/aromatics/resins/asphaltenes fraction amounts). − We wanted a relevant process that could be performed on these mixtures that would produce a quantifiable and molecular outputwhere we would aim to develop a ML approach that could predict such an output. The specific character of asphaltenes presents an additional challenge for obtaining data of this nature.…”