Streamlined method for estimating hydrocarbon fuel properties using a compact, economical Fourier Transformation Infrared-based (FTIR) analyzer

Stanford researchers have developed a streamlined method for simultaneously estimating a broad range of hydrocarbon fuel physical and chemical properties for a wide range of fuels. These properties are estimated with vapor-phase mid-infrared spectra (measured using a commercially-accessible FTIR spectrometer) and machine learning models.
This invention could be used to provide fast, accurate, and economical estimations of hydrocarbon fuels' properties with small amount of sample during the early stage of alternative fuel development. It has the potential to expedite the process of alternative fuel certification, which is one major hurdle for market acceleration of the global alternative fuel and hybrid vehicle industry.

Stage of Research

• Proof-of-concept
• The team demonstrated that mid-infrared spectra together with regularized linear models trained with a dataset consisting of 64 hydrocarbon fuels can accurately estimate 15 physical and chemical properties
• Ongoing research to extend wavelength range and to analyze more parameters