A method for the rapid, accurate prediction of the physical properties of middle distillate fuels from LC-¹H NMR derived data

Abstract

A method has been developed whereby various physical properties of middle distillate fuels may be rapidly and accurately calculated by a group property approach from data obtained from a directly coupled Liquid Chromatograph -⁻¹H Nuclear Magnetic Resonance Spectrometer (LC-⁻¹H NMR). The physical properties include cetane number, cetane index, density, specific gravity, pour point, flash point, viscosity, filterability, heat of combustion, cloud point, volume percent aromatics, residual carbon content, and initial, 10%, 50%, 90%, and end boiling points. These property predictions have accuracies approaching the error for measurement of the experimental physical property and require less than two hours analysis time per fuel. An interface was developed between the NMR spectrometer and a personal computer to aid in automation of the LC-⁻¹H NMR data collection and to perform off-line analysis of the LC·⁻¹H NMR data. This interface and all associated software is described.

Also presented is a series of model compound studies in which the physical properties of pure hydrocarbons (i.e., alkanes, monocyclic and dicyclic aromatics) were predicted by a similar group property approach.