PCA As Tool For Intelligent Ultrafiltration For Reverse Osmosis Seawater Desalination Pretreatment

Abstract A novel fouling monitoring methodology based on principal component analysis (PCA) has been validated using transmembrane pressure (TMP) data of a pilot-scale pressurized ultrafiltration (UF) system operated with seawater. The evolution of membrane fouling was investigated to determine its relation to the used cleaning strategy on the one hand and the quality of the raw seawater on the other hand. The developed models showed that in terms of cleaning efficiency there are no significant differences between the standard and optimized backwashing protocols that were employed. This confirms the hypothesis of being able to use the optimized operation in a sustainable manner and benefit from lower cleaning frequencies. In addition, it has been demonstrated that the use of PCA as a monitoring technique to detect abnormal fouling behaviour is a robust tool. By using PCA, decisions on cleaning sequences or frequencies could be taken dynamically instead of running the system with fixed cycles.