(484a) Modeling of Long-Time Flux Decay In Cross-Flow Filtration

Recent work at Pacific Northwest National Laboratory has been concerned with understanding how cross-flow filter flux decays with time.  Understanding the cause(s) of flux decay is important for assessing long-term behavior of this separation technology, especially when used in challenging processing environments e.g. nuclear waste treatment.  Fouling in the filter is a complex issue that is affected by many parameters: the chemistry of the solution being filtered, particle size and morphology, operating conditions, and cleaning methods.  Based on filtration work performed at PNNL over the past several years, it is clear that a portion of the fouling can be “reversed” by back-pulsing the filter and another portion is irreversible (usually only rectified via cleaning protocols).  However, the sheer number of parameters requiring consideration complicates effective modeling of both the long-time behavior of the cross-flow filter and its response to back-pulses. 

To combat this, simple models were developed from first principles based on different suspected fouling mechanisms.  The models were tested against long-term data (100-hour tests at a constant solids concentration) collected with PNNL’s bench-scale filtration system.  The flux was found to be well described by a cake mechanism model at intermediate times (starting at approximately 3 hours elapsed time); at longer times (approximately greater than 20 hours of filtration time) the flux behavior transitions to an intermediate mechanism behavior.  The decay constant for these models can be predicted using a single dimensionless parameter, the ratio of the transmembrane pressure to the inertial pressure.  Additional testing suggested that the filter could be conditioned by back-pulsing.  These observations were used to build a comprehensive model to assist in determining a back-pulse strategy for cross-flow filters that will experience long-term decays in the filter flux.