Prediction of freezing point of jet fuel compounds using group contribution method

Abstract

Freezing point (TF) is the temperature at which a liquid substance turns into a solid when cooled and this can be determined experimentally or by modelling. For aviation fuels, the freezing point test is relevant as impeding fuel flow may have disastrous results on planes, such as interference with fuel atomization. The freezing point of jet fuel can negatively be affected due to certain components found in the fuel as blend. In this study, a modelling method called Group Contribution method (GC) is used to predict the freezing point of jet fuel. The GC approach was tested and verified using different chemical data set. This research consisted of predicting the freezing point of 232 compounds using the first order-groups, second order-groups and third order-groups identified from the molecular structure and proving the validity of the model using 20 test data set of compounds. The regression using linear regression curve fitting using MATLAB was performed to obtain the contribution values for each group. The statistical analysis giving a standard deviation (SD) of 22.22 K, an average absolute error (AAE) of 15.62 K, an average relative error (ARE) of 7.23 % and a correlation coefficient, R2 of 0.8829. The model was tested and verified to be an appropriate method to predict the freezing point of jet fuel liquid compounds.