Abstract
Evanescent field absorption spectroscopy in the infrared range of coiled fiber-optic sensor has been applied to determine the low-level water content in ethanol. Principal component regression and classical least square models have been utilized to build the calibration model and predict the water concentrations. The standard errors of predictions of water concentrations in ethanol were 3.16% and 0.42% respectively. Some methods to improve the accuracy of predicted water concentration in ethanol were suggested. The predicted concentration of water is acceptably accurate and cost-effective. The study demonstrates that the coiled fiber-optic sensor based on evanescent absorption spectroscopy is a feasible technology for prediction of the low-level water content in bio-ethanol and other industries in both on-line and remote situation.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
