Operation and performance analysis of direct hollow fiber nanofiltration: A pilot study at IJsselmeer

• Published in: Separation and purification technology Vol. 349; p. 127786
• Main Authors: Ophorst, Marleen, Grooth, Joris de, Heijman, Sebastiaan G.J., Vaudevire, Elisabeth M.H., Jafari, Morez
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 03.12.2024
• Subjects: Drinking water treatment; Fouling potential; Hollow fiber nanofiltration; PFAS; Pilot; Drinking water treatment; Pilot; Fouling potential; Hollow fiber nanofiltration; PFAS
• ISSN: 1383-5866
• DOI: 10.1016/j.seppur.2024.127786

•The dNF40 membrane exhibited minimal fouling when operated with raw surface water.•Retention for various contaminants, including ions, NOM and OMPs, were investigated.•An increase in permeability was observed over the course of long-term operation.•Economic feasibility of a full-scale dNF40 membrane treatment plant was evaluated. This study investigated the performance of direct hollow fiber nanofiltration (dNF40) membranes from NX Filtration BV on a pilot scale for the treatment of pre-treated IJsselmeer water from Waterwinstation Prinses Juliana (WPJ) in Andijk, as well as the direct treatment of raw IJsselmeer water. The objective was to evaluate the long-term fouling potential and the retention of ions and natural organic matter (NOM) using both WPJ pre-treated IJsselmeer water and raw IJsselmeer water. Additionally, the rejection of organic micropollutants (OMPs) under artificially elevated conditions, referred to as ‘spiked solution’, using WPJ pre-treated IJsselmeer water was investigated. Limited to no fouling was observed on the dNF40 membrane during stable operation when treating both WPJ pre-treated IJsselmeer water and raw IJsselmeer water, even under changing process conditions. NOM removal consistently exceeded 90% regardless of process conditions or water type. The retention of per- and polyfluoroalkyl substances (PFAS) was above 80%, with even higher retention observed for higher molecular weight values. Low molecular weight pharmaceuticals, all below the molecular weight cut-off (MWCO) of the dNF40 membrane (400 Da), exhibited approximately 30% retention. The dNF40 membrane showed better retention of negatively charged pharmaceuticals in the spiked solution compared to positively charged and neutral pharmaceuticals. A total cost of ownership (TCO) analysis unveiled that operational expenditures (OPEX) were three times higher than capital expenditures (CAPEX) for a 5-stage full-scale dNF40 system. Among the components, membrane replacement costs constituted the majority of OPEX (68%), followed by energy costs (31%) and chemical costs (<1%). Overall, the study showcased the suitability of the dNF40 membranes for treating IJsselmeer water, achieving effective removal of NOM and PFAS.