Brackish water recovery from reactive dyeing wastewater via ultrafiltration

Abstract

In this study, brackish water recovery from a real reactive dyeing textile effluent for reuse of water and salt was investigated. With the aim of improving the overall retention of species that are partly retained in a single stage, three different two-stage ultrafiltration scenarios were considered: 5 kDa followed by 5 kDa, 5 kDa followed by 2 kDa and 2 kDa followed by 2 kDa. The 2 kDa + 2 kDa scenario reduced the total organic carbon (TOC) from 239 ± 9 mg/L to 41 ± 7 mg/L, at 2 bar of transmembrane pressure (TMP) until a volume reduction factor (VRF) of 2.5, while color was reduced to less than 1% of the feed. The first stage 2 kDa at 4 bar experienced slight flux decline, however, the fouling was fully reversible just with physical cleaning. With this scenario, when the VRF was increased to 10 to increase the water recovery and TMP was increased to 4 bar to increase the permeate flux, the permeate quality remained similar, even when the pollutant load in the feed stream was doubled. With the 5 kDa + 2 kDa scenario, while TOC and color removals were somewhat less, the average first stage flux was about four times higher than that of the first stage 2 kDa filtration, although flux decline and partial fouling irreversibility were still observed. On the other side, the 5 kDa + 5 kDa scenario was the poorest in performance as 5 kDa membrane showed slightly lower retention of color and TOC with higher flux decline and poorer fouling reversibility; although the permeate flux was high. Considering the advantages of high retention and high flux, the permeate samples from both 2 kDa + 2 kDa and 5 kDa + 2 kDa scenarios were tested in fabric dyeing and found to be successful even with a color that is different than the color of the wastewater. These findings suggest that the proposed treatment scenario to recover salty water for reuse can be applicable for wastewater of changing characteristics, which is typical for textile dyeing processes.