Characterization of hydraulically reversible and irreversible fouling species in ultrafiltration drinking water treatment systems using fluorescence EEM and LC–OCD measurements

Abstract

The application of the fluorescence excitation-emission matrix (EEM) approach and liquid chromatography–organic carbon detection (LC–OCD) analysis for the characterization of hydraulically reversible and irreversible fouling species, extracted from hollow fiber ultrafiltration (UF) membranes used in drinking water treatment, was demonstrated. Hydraulically reversible and irreversible fouling species were extracted from two pilot UF membrane systems operated in parallel with lake water as the feed. Two membrane cleaning protocols, hydraulic- and chemical-based (NaOCl and citric acid) cleaning, were considered. Colloidal/particulate matter together with protein-like and metal species in water appeared to contribute to the formation of a hydraulically removable fouling layer on the membranes. Hydraulically irreversible fouling, in contrast, was impacted considerably by humic substances (HS) and protein-like matter. The formation of an irreversible fouling layer was also likely influenced by interactions between the colloidal/particulate matter and metal species together with HS and protein-like matter. LC–OCD analysis revealed the presence of predominant levels of lower molecular weight HS-like matter – compared to the HS-like matter commonly present in lake water – in the citric acid extracted foulant fraction. The permeability loss due to hydraulically irreversible UF fouling was considerably greater than the permeability loss due to hydraulically reversible UF fouling. A permanent permeability loss (∼25–35% of the initial permeability) was present even after the application of considerably strong chemical cleaning protocols on both pilot systems. This study indicated that the fluorescence EEM approach can be applied for monitoring and characterization of membrane cleaning procedures and as a potential diagnostic tool for assessing the effectiveness of hydraulic- and chemical-based cleaning protocols employed in UF drinking water treatment operations using rapid off-line measurements. On the other hand, since the LC–OCD analysis technique is a comparatively time consuming method, it may be used for verification of the fluorescence EEM-based results of the foulant fractions.

Composition and reversibility of fouling on low-pressure membranes in the filtration of coagulated water: insights into organic fractions behaviour

“…Consequently, there seems to be no need to apply any CEB under these filtration conditions, with the consequent lower reagent consumption and extension of membrane life. The higher detachment of BP by BW than other 10 organic fractions was in accordance with other studies [20,22]. Differences against other studies appear, however, in the percentage detachments.…”

“…3 is in agreement with Peiris et al [20], who carried out UF of surface water under similar conditions (permeation duration of 30 min and BW duration of ca 2 min) and 30 reported recoveries of the initial permeability of <10% when BWs were applied and of 65-75% when chemical cleaning based on NaClO and citric acid in sequence was employed. In contrast, when treating surface water by UF (permeation duration of 30 min 35 and BW duration of 10 min), Wray et al [21] found that most of the total fouling observed (>70%) was hydraulically reversible.…”

“…The preferential washing out of the BP fraction was likely due to its size relative to that of the membrane pores: organic substances much larger than the 70 membrane pores led to the formation of a cake weakly bound to the membrane and thus more readily washed out [21,22,26], while lighter fractions such as HS or microflocs of Al and Fe can cause pore blocking or build-up a denser and tight cake layer more closely 75 adhered to the membrane surface and thus less readily detached from it by BW [14,16,20]. Differences between BW and BW(+CEB) were not appreciable, NaOH under CEB conditions.…”

“…Some of these studies have been carried out with synthetic solutions [12,13] or organic model compounds [14][15][16] often at high con-5 centrations compared to those in real waters providing results that are not always comparable to practical situations, while some others have been carried out with real waters of different origin such as surface water [2,14,[17][18][19][20][21], wastewater [11,[22][23][24][25] and seawater 10 [26,27], each with different fouling behaviour potentials. Besides, most of these works determine fouling formation in terms of membrane hydraulic resistance using transmembrane pressures (TMP) and fluxes based on Darcy's Law [11,15,23,24], but do not 15 quantify the amount of foulant accumulated on the membrane.…”

A membrane combined process to cope with algae blooms in water

“…The eluted water after alkaline cleaning showed two fluorescence areas, region D (Ex/Em = 330/400 nm) and region C (Ex/Em = 240/400 nm). In contrast, the organics from acidic cleaning only had one fluorescence area (Ex/Em = 250/350 nm), which was marked as region F. As indicated in previous findings, region D is considered as a fluorescence response of the humic-like organics, while region C is believed to represent the hydrophobic and protein-like organics [46]. Although alkaline cleaning water had high concentrations of b a N-HPI organics (Fig.…”