Skip to Main Content
In this paper, the measurement of process streams and effluents from the sugar-ethanol industry using an optical fiber sensor based on the Fresnel reflection principle is reported. Firstly, binary sucrose-water and ethanol-water solutions with predetermined concentrations were measured for calibration purposes. Secondly, the coproducts from various processing stages were analyzed in order to identify the sucrose or ethanol content. The measured data were processed by an artificial neural network model, which correlated the reflected intensity values to the sample concentration. The absolute error was calculated by a comparison between the nominal concentration values obtained by the plant laboratory analysis and the sensor response, yielding errors ≤ 3& wt% and ≤ 5.1 vol% for the sucrose and ethanol contents, respectively. The fiber sensor has the potential to provide reliable results even for samples with more complex compositions than pure sucrose or ethanol solutions, with perspectives of application on the several stages of the plant facility.