WaterSubject

Domestic wastewater offers the potential to meet some of the expected growth in water demand in the future. Greywater originates from all household sources other than the toilet, urinal or bidet, and hence does not usually contain faecal matter; greywater constitutes almost half of the volume of domestic wastewater. Foul water, on the other hand, includes all household wastewater entering the sewer system.

Processes whereby membranes are coupled with biological treatment are known to produce high-quality effluent that is free from bacteria, viruses and suspended solids. The biological aerated filter (BAF) provides a high-intensity wastewater treatment that produces effluent quality that cannot be achieved by conventional biological processes, such as activated sludge and trickling filters. However, a membrane stage such as microfiltration (MF) or ultrafiltration (UF) is required to provide a complete barrier to bacteria and viruses, which are not rejected by the BAF. The high-quality effluent from such systems then facilitates a wide range of potential water-reuse applications, including nonpotable domestic applications and irrigation. In addition, UF and MF membranes are an ideal pretreatment for reverse osmosis.

Fouling remains the major constraint in the application of membranes to domestic wastewater treatment. It leads to a decline in membrane permeability, making more frequent membrane cleaning and replacement necessary and thus increasing operating costs and decreasing plant output. Previous studies have revealed that productivity of UF membranes to declines rapidly with increasing operating flux, attributable to enhanced solids and organic loading and compaction of the fouling cake layer on the membrane surface.

This paper is an extension of a previously reported study based on the same apparatus, in which permanent fouling behaviour was described empirically in terms of the effective membrane area A_e at a given time available for filtering:

(1) where ?P and ?P₀ are the transmembrane pressure values initially and at time t, and A₀ and A_e are the corresponding values for the available membrane area.

Membrane fouling is an inherently complex and variable process. By applying statistical techniques, the influence of each parameter and the level of uncertainty can be quantified. Results are examined statistically so as to produce numerical relationships between performance and key process conditions appropriately qualified through calculation of the innate variability and uncertainty.

The domestic wastewaters used comprised a synthetic greywater, real greywater (from Cranfield University’s Feddon House greywater collection facility) and settled sewage (from the university’s own sewage treatment works). The synthetic greywater was prepared according to a recipe developed in previous greywater studies. These waters were all pretreated using a BAF before UF using capillary tubes.

The test rig, feedwater composition and all other experimental details have been reported previously. Trials were conducted so as to investigate the impact of flux (J), backwash cycle time (t_b) and backwash duration (t_d) on overall productivity for the three wastewater types. Performance was quantified in terms of both the pressure increase for constant flux operation and the specific throughput S in mbar:

(2) where A₀ is the membrane area, and ?V and ?P are the volume and pressure change, respectively, over the course of some representative discrete time period, such as the cleaning cycle time t_c. Trials were repeated at least three times for each process setting, and the membrane was chemically cleaned between trials.