Prediction of the oxidative stability of biodiesel based on the olfactory profile

Biodiesel (BD), especially biofuel derived from waste cooking oil (WCO), is a sustainable alternative to fossil fuel but the commercial viability of the product is hampered by its susceptibility to oxidation. Traditional methods for assessing oxidative stability, such as the Rancimat method, are expensive, time-consuming and impractical for real-time monitoring. This study presents a promising and low-cost approach to predicting the oxidative stability of WCO-derived BD using an electronic nose (e-nose) combined with stochastic modeling. BD was synthesized by transesterification with ethanol using a heterogeneous catalyst, and the effects of a natural antioxidant (grape pomace extract, GPE) and a synthetic antioxidant (butylated hydroxytoluene, BHT) on the stability of the biofuel were investigated after 30 and 60 days storage. Stochastic modeling was highly effective in describing olfactory profiles, with most of the 32 e-nose sensors showing coefficients of determination (R 2 ) greater than 70 %. A multiple linear regression model was developed from the stochastic parameters that successfully predicted oxidative stability as demonstrated by a R 2 of 75.8 % based on average readings. The results demonstrated that, under the evaluated conditions, the combination of GPE (1000 ppm) and BHT (200 ppm) produced a significant synergistic effect, increasing oxidative stability by approximately 50 % and outperforming the antioxidants applied individually (GPE ∼ 6 % and BHT ∼ 34 %). The proposed methodology is a robust, fast, and accessible tool for the quality control of BD because it allows the differentiation between the effects of additives and aging, thereby presenting great potential for the optimization of BD production and storage.