Analysis of mixed ratio of waste cooking-oils in biodiesel production process using near-infrared spectroscopy

Waste cooking-oil is one of the raw materials for producing biodiesel (BD). When the waste cooking-oils are collected for the BD production, they are mixed with various kinds. During the BD production, carefully selected kinds, optimized amount of alcohol and catalyst are mixed with the waste cooking-oil mixture for the maximum yield. However, those conditions are highly dependent on the kinds and mixed ratio of waste cooking-oils. In this research, we aim to develop a system measuring the mixed ratio of raw materials of waste cooking-oil through near-infrared spectroscopy. Two to four kinds of raw materials of waste cooking-oil (soybean oil, canola oil, sunflower seed oil, and olive oil) were mixed with combinations of a different ratio. The NIR (Near-infrared) absorbance of the mixture was measured by placing between tungsten-halogen light source and the NIR spectrometer (898.677 - 2132.65nm). These spectra were used to execute PLSR (Partial Least Squares Regression) on each raw material of waste cooking-oil concentration. As a result, 28 PLS models were generated, showing high linearity and low prediction error of the mean values of the models of R^2_cal = 0.987, R^2_val = 0.979, RMSEC = 0.203, RMSEV = 0.250. It was confirmed that the absorption spectrum of mixed waste cooking-oil was estimated and the content of each raw material could be measured. It is expected that applying near-infrared spectrometer and PLS model to BD production process for the production of BD using waste cooking-oil will greatly contribute to the improvement of BD yield.