Techno-economic analysis of textile dye bath wastewater treatment by integrated membrane processes under the zero liquid discharge approach

Abstract The present study was intended to determine the usefulness of the zero liquid discharge (ZLD) approach in treating textile dye bath wastewater via integrated membrane processes. These processes included various combinations of ultrafiltration (UF), loose nanofiltration (NF1), tight nanofiltration (NF2) and reverse osmosis (RO). The technological and economic performance of four scenarios (UF/NF2 (S (I)), NF1/NF2 (S (II)), NF1/RO (S (III)) and UF/NF2/RO (S (IV))), each followed by membrane distillation (MD), were comparatively analysed. The factors considered were capital and operating costs, revenues, benefit/cost (B/C) ratios and pay-back times. The last two scenarios, those including RO, provided the best effluent quality at the end of the pressure-driven membrane filtration process. However, these scenarios were not the most economical options because of the decreased benefits of recycling soda ash and NaCl obtained from the treated wastewater. The unit treatment costs of the scenarios were 1.37, 1.38, 2.16 and 2.01 $/m3 of influent for S (I), S (II), S (III) and S (IV), respectively, with return periods of 0.87, 0.91, 2.07 and 1.51 years. The best technological and economic performance was estimated for S (I) and S (II), as indicated by the B/C ratios 3.58 and 3.55, respectively. At the break-even point of 15 years, S (I) required an incineration cost of 77.5 $/m3 of MD concentrate, whereas the corresponding costs associated with the other three scenarios were estimated to be 75.7, 43.0 and 35.5 $/m3 of MD concentrate, respectively. It was finally concluded that the ZLD approach is most useful in processing textile dye bath waste and recycling treated wastewater. This process is both technically feasible and economically viable, as indicated by the high B/C ratio. However, the unit incineration cost associated with the concentrate would limit the return period of the investment, and the success of this method would also depend principally on the real volume of MD processed due to its large contribution (70–90%) to the B/C ratio.