A Group Contribution Method for Estimating Cetane and Octane Numbers

Kubic, William Louis¹ 

[1] Los Alamos National Lab. (LANL), Los Alamos, NM (United States). Process Modeling and Analysis Group

关键词: Energy Sciences;    Inorganic and Physical Chemistry;    Organic Chemistry;    Cetane Number;    Research Octane Number;    Motor Octane Number;    Group Contribution Method;    Artificial Neural Networks;   

DOI  :  10.2172/1291241
RP-ID  :  LA-UR--16-25529
PID  :  OSTI ID: 1291241

学科分类：燃料技术

美国|英语

来源: SciTech Connect

【 摘 要 】

Much of the research on advanced biofuels is devoted to the study of novel chemical pathways for converting nonfood biomass into liquid fuels that can be blended with existing transportation fuels. Many compounds under consideration are not found in the existing fuel supplies. Often, the physical properties needed to assess the viability of a potential biofuel are not available. The only reliable information available may be the molecular structure. Group contribution methods for estimating physical properties from molecular structure have been used for more than 60 years. The most common application is estimation of thermodynamic properties. More recently, group contribution methods have been developed for estimating rate dependent properties including cetane and octane numbers. Often, published group contribution methods are limited in terms of types of function groups and range of applicability. In this study, a new, broadly-applicable group contribution method based on an artificial neural network was developed to estimate cetane number research octane number, and motor octane numbers of hydrocarbons and oxygenated hydrocarbons. The new method is more accurate over a greater range molecular weights and structural complexity than existing group contribution methods for estimating cetane and octane numbers.