Modeling, optimization and control in dead-end ultrafiltration

Ultra filtration (UF) is increasingly used as a complete- or intermediate surface water purification technique. Ultra filtration membranes have a high selectivity, are easy to scale-up and have become economically attractive during the last 15 years. However, the buildup of fouling during filtration severely influences the performance of a UF membrane. For this reason frequent cleaning of the UF membrane is required. In the short term, the membrane is cleaned by means of backwashing, and in the long term the membrane is cleaned with cleaning chemicals. Although backwashing and chemical cleaning are useful methods to prevent fouling, the execution of such procedures are still based on pilot plant studies and rules of thumb. The expectation is that systematic modeling; optimization and control will significantly reduce operational costs and increase overall controllability of the UF process. In this contribution the formulation of a hierarchical modeling, optimization and control framework will be discussed that can be used to optimize the overall UF process. There will be special attention for mathematical modeling and optimization of membrane filtration, backwashing and chemical cleaning. Noted is further that the proposed models, optimization routines and control algorithm were tested and validated experimentally, on lab scale as well as pilot scale.