Probabilistic Mean Quantitative Structure–Property Relationship Modeling of Jet Fuel Properties

Abstract

We present a novel probabilistic mean quantitative structure–property relationship (M-QSPR) method for the prediction of jet fuel properties considering two-dimensional gas chromatography measurements. Fuels are represented as one mean pseudo-structure that is inferred by a weighted average over structures of 1866 molecules that could be present in the individual fuel. The method allows training of models on both data of pure components and of fuels and does not require mixing rules for the calculation of the bulk property. This drastically increases the number of available training data and allows the direct learning of the mixing behavior. For the modeling, we use a Monte-Carlo dropout neural network, a probabilistic machine learning algorithm, that estimates prediction uncertainties due to possible unidentified isomers and dissimilarity of training and test data. Models are developed to predict the freezing point, flash point, net heat of combustion, and temperature-dependent properties such as density, viscosity, and surface tension. We investigate the effect of the presence of fuels in the training data on the predictions for up to 82 conventional fuels and 50 synthetic fuels. The results of the predictions are compared on three metrics that quantify accuracy, precision, and reliability. These metrics allow a comprehensive estimation of the predictive capability of the models. For the prediction of density, surface tension, and net heat of combustion, the M-QSPR method yields highly accurate results even without the presence of fuels in the training data. For properties with nonlinear behavior over temperature and complex fuel component interactions, like viscosity and freezing point, the presence of fuels in the training data was found to be essential for the method.

Screening and evaluation of biomass upgrading strategies for sustainable transportation fuel production with biomass-derived volatile fatty acids

“…The suitability of upgraded alcohols and alkanes for blending depends, in turn, on the fuel properties of individual upgraded bioblendstock components, properties of fuel mixtures, and application-specific limitations on fuel properties that ensure engine safety and operability. We identified seven fuel properties: boiling point (T90, BP), flash point (FP), energy density (lower heating value, LHV), viscosity (at 40°C for diesel, −40°C for jet, and 25°C for gasoline; μ), melting point (cloud point for diesel, MP), water solubility (WS), and ignition quality (cetane number (CN) for diesel and jet; research octane number (RON) for gasoline), as broadly important for diesel, 35 jet, 2 , 19 , 36 , 37 and gasoline 38 applications and gauged operability of fuels based on conformation to limits on these properties. Individual component molecular fuel properties are listed in our GitHub repository ( Note S2 ), while the mixture property calculation formulas are shown in Table S1 .…”

Theoretical modeling and machine learning-based data processing workflows in comprehensive two-dimensional gas chromatography—A review

“…4). However, it is important to acknowledge the potential applicability of other ML-based algorithms, including linear discriminant analysis (LDA) [20,105], Monte-Carlo neural network (MCNN) [108], cluster resolution (CR) [109], and even customized ML software [110].…”

Comprehensive Two-Dimensional Gas Chromatography: A Universal Method for Composition-Based Prediction of Emission Characteristics of Complex Fuels

“…The mandatory conversion process toward sustainable combustion does not change this discrepancy but rather expands the complexity of technical fuels through additional chemical structures through novel alternative production pathways such as biofuels − and Power-to-X, − including solar production routes. Starting from the chemical composition, application of the molecular knowledge enables the determination of numerous fuel properties such as density, freezing point, flash point, heat of combustion, surface tension, and viscosity. , As a consequence, a very detailed understanding of the composition of complex fuel mixtures becomes the crucial factor in linking molecules to combustion processes. Detailed compositional determination can be achieved by comprehensive two-dimensional gas chromatography (GCxGC), playing an important role in the analysis of petroleum-derived products, fossil fuels, and increasingly synthetic fuels and fuels from bio-based sources. − …”