Diagnostic assessment of membrane health and performance by s-transform of flux data

Abstract

The time-frequency analyses are extensively used in electrical engineering, medical imaging, image processing, and acoustics. Analysis of system response in relation to both frequency and amplitude provides diagnostic insight as to when the system experiences significant stress resulting in shifts from its previous performance. S-transformation allows identification and separation of system responses to different stress factors that occur at different times. S-transform of time dependent data generates visual images (i.e., contours) which can be interpreted based on the physical characteristics of the system. In this study, S-transformation was used to evaluate membrane stress due to fouling during cross flow filtration with membranes with moleculer cut off (MWCO) ranging from 5K to 30 KDa. Flux data from cross flow filtration experiments were analyzed to visualize membrane health and levels of stress due to fouling were analyzed. Characteristic changes in flux profile were interpreted in relation to membrane specific fouling mechanisms of the membranes in relation to MWCO. © 2011 ASCE. © 2011 Elsevier B.V., All rights reserved.