Reproducibility and applicability of the flux step test for a hollow fibre membrane bioreactor

Results from flux step tests conducted on a pilot-scale hollow fibre immersed membrane bioreactor plant (HF iMBR) are reported. Trials combined a classic flux-step test, each step incorporating six backflush cycles, with an extended (15-h) test to evaluate trends in fouling. Both fouling rate, i.e. the rate of increase in transmembrane pressure (dTMP/dt) at a constant flux, and mean permeability were calculated and reproducibility assessed through the standard deviation. Applied nominal fluxes between 5 and 35 L m−2 h−1 (LMH) were employed for the flux-step trials, and all experiments were (i) triplicated through simultaneous application across three central membrane elements in the seven-element membrane module, and (ii) repeated four times. Reproducibility was quantified as standard deviation (SD), and assessed with reference to both the mean permeability across the backflush cycle and the irreversible fouling rate across the three modules – the latter both for the short-term and long-term trials.Results indicated greatly reduced fouling (by 80–88% with respect to fouling rate) for the longer-term study compared with the flux-step data. Fouling rate SD values were slightly lower for the flux step data at 20–27% for the four highest flux steps compared with 30% for the longer trial conducted at 20 LMH, based on quadruplicated data. Mean permeability data, on the other hand were much more reproducible, with SD values below 8%. It was concluded that significant variations in hydrodynamics can arise across a module comprising a number of elements, accounting for a 20–40% change in fouling rate. However, this cannot account for the differences recorded between the short-term and long-term fouling measurements, which are a facet of the flux step test conditions.

► Fouling of three individual membrane elements in pilot-scale MBR module investigated.
► Fouling rate recorded for 15 h trial 80–88% lower than that from 1 h flux step test.
► Standard deviation (SD) of ∼30% recorded for quadruplicated fouling rate measurements.
► SD of <8% measured for mean permeability with supracritical flux indicated throughout.
► Flux step test trend atypical, yielding overly conservative design flux guide value.