Hydrocarbon analysis and control by raman spectroscopy

摘要

Benzene and substituted aromatic hydrocarbons can be predicted within +/-0.31% vol or better, using Raman NIR spectroscopy and multivariate analysis, with optional fiberoptics multistreaming, preferably with Partial Least Squares regression analysis. The resulting signal can be used to control concentration of such compounds in product to desired levels.