Predicting chlorine demand for fresh produce washing to ensure disinfection efficacy and minimize production of disinfection by-products
The University of Georgia
Professor Yen-Con Hung
This study was conducted to develop models capable of predicting chlorine demand of different fresh and fresh-cut produce during produce postharvest washing treatment. Simulated produce wash water with different chemical oxygen demand (COD) were prepared using ten different fresh fruits and vegetables. Water quality parameters including pH, oxidation reduction potential (ORP), ultraviolet absorbance measured at 254nm (UV254), COD, turbidity, total protein content, total phenolic content and color difference between wash water and deionized water (ΔE) were measured. In addition, sodium hypochlorite (NaOCl) was added to wash water to determine the chlorine demand. Correlations between wash water quality and chlorine demand were analyzed. The results shows that UV254 had the highest correlation coefficient with chlorine demand (R=0.77). Further analysis of scatter plot between chlorine demand and UV254 shows two clusters exist: clusters for produce with high phenolic and low phenolic content. Based on chlorination reaction mechanism, the phenolic-to-protein/ΔE ratio (PPC) was created to determine the cluster of each sample. Equations for predicting chlorine demand were developed. The prediction models were further validated using both same type of produce but at different COD and different produce than previously used, with prediction error of 11.3 and 8.16, respectively. The validation results indicate that models have consistent prediction capacity of chlorine demand for different types of fresh produce.