Volume change of CO2 + long-chain liquid n-alkane (with n ≥ 14) mixtures under geological conditions: An observed abrupt change near the critical point of CO2
Description
This study used a high pressure optical cell (HPOC) combined with a heating/cooling stage, a pressure device, and a laser Raman spectrometer to investigate the volume expansion of CO2 + tetradecane and CO2 + hexadecane mixtures under geological conditions. Abrupt changes in volume expansion were observed near the CO2 critical point, which is assumed to be ubiquitous in CO2 + long-chain alkane mixtures with ≥ 14 carbon alkanes. This is due to the transformation of the intermolecular structure of the mixture, which inhibits the volume expansion of CO2 + long-chain alkanes at lower temperatures. In addition, four machine learning (ML) algorithms, including backpropagation neural network (BPNN), K-nearest neighbors (KNN), random forest (RF), and support vector regression (SVR), were applied to predict the volume expansion factor. The BPNN model provided much higher robustness and accuracy with correlation coefficient (R2) > 0.98 and mean square error (MSE) < 0.03, which indicates a strong agreement between the predicted result and matching experimental data. We concluded that a predictive framework built using ML could provide rapid predictions of the volume change curve or curved surface of CO2 + organic mixtures under various temperature and pressure conditions.