Ultrafiltration (UF) or microfiltration is often employed
during
potable reuse for reverse osmosis pretreatment, and their own fouling
from organics and colloids present in the secondary wastewater effluent
remains an unresolved issue. Herein, we compared the fouling of a
hollow fiber UF membrane over multiple filtration–backwashing
cycles caused by two different secondary-treated effluents having
similar concentrations of effluent organic matter (EfOM) and dissolved
constituents but substantially different turbidity, total suspended
solids (TSS), and particle size distributions. Contrary to expectations,
significantly worse fouling was caused by the less turbid (low TSS)
wastewater, which was explained by the presence of particles that
were smaller than the membrane pores. In this case, pore plugging
was the dominant fouling mechanism, which was mechanistically quantified
as “standard blocking,” i.e., colloid deposition inside
the pores. On the other hand, fouling was negligible for the more
turbid (high TSS) wastewater that contained particles larger than
the pores. In this case, surface accumulation of colloids blocked
pores, which was mechanistically quantified as “intermediate
blocking.” Hence, the location of particle deposition was a
crucial determinant of ultrafilter productivity when colloids combined
with EfOM to (ir)reversibly foul membranes operating on real-world
secondary effluents. Additionally, pretreatment by electrocoagulation
and conventional coagulation mitigated (ir)reversible fouling by creating
large flocs that deposited on the membrane surface, (i) allowing facile
backwashing (i.e., reduced irreversible fouling) and (ii) forming
a more permeable cake layer (i.e., reduced reversible fouling). Hence,
UF productivity was primarily determined by colloid size relative
to pore size in conjunction with secondary contributions from EfOM.
Importantly, internal pore fouling within the membrane’s polymer
matrix was more difficult to alleviate by hydraulic backwashing than
by surface deposition.