The objective of this study was to evaluate the temporal behavior of post filter
adsorbers employing different types of granular activated carbon (GAC) media in the
removal of precursors of chlorinated organics and in the formation of disinfection
byproducts. The experimental setup was four pilot scale adsorbers, two units with
mineral GAC media (CALGON Filtrasorb 300) and another two units with vegetal
GAC media (CARBOMAFRA CAG 119). The four adsorbers (F1 through F4) were
operated in parallel, with the same empty bed contact time (EBCT) of about 20 minutes.
Two of them (F1-CALGON and F2-CARBOMAFRA) were fed with filtered water
from a Alto da Boa Vista Water Treatment Plant (ABVWTP) which was ozonated prior
to feeding, in order to observe any effects of ozonation on the different adsorbers. The
other two adsorbers (F3-CALGON and F4-CARBOMAFRA) were fed with filtered
water only. The ozonation system was composed of four column shaped contactors in
series. Each column had a hydraulic detention time of 4 minutes. Ozone dosages were
applied in the first and second columns of the series. The total ozone dosage was in the
range 0.2 through 1.0 mg O3/L and was equally split between columns 1 and 2.
Evaluations of precursor removal and disinfection byproduct formation were by means
of COT, UV-254 nm determinations and THM formation potential, respectively. The
filtered water was from the conventional ABV WTP, whose raw water source is from a
highly eutrophic reservoir with a long detention time. COT and UV-254 values for the
ozonation system feed water and for adsorbers F3 and F4 were in the ranges 2.0 to 2.5
mg C/L and 0.02 to 0.04 cm-1, respectively. It was observed that ozonation did not
affect significantly the COT and UV-254 values. During the first thirty days of post
filter adsorber operation, COT removal efficiencies were around 80 % for F1 and F3
(with CALGON GAC media) and around 40 % for F2 and F4 (with CARBOMAFRA
GAC media). With sixty days of operation, COT removal efficiencies decreased to 40
% and 20%, respectively and stayed that way during eight months of operation. The
mineral GAC media performed better regarding COT and UV-254 removals than the
vegetal GAC media. On the other hand, no significant difference in effluent quality was
observed regarding adsorbers with the same GAC media type receiving ozonated and non-ozonated filtered water. Thus ozonation did not seem to affect the adsorbing media.
Regarding THM formation, adsorbers F1 and F3 presented a better performance than
their counterparts (F2 and F4), the same pattern observed regarding COT and U-254
removals. During the first thirty days of adsorber operation, THM reduction efficiencies
were about 85 % for F1 and F3 and 50 % for F2 and F4. With eight months of
operation, such numbers stabilized around 50 % and 30 %, respectively, with adsorber
behavior similar to what happened with regard to COT removal. Continuous, long term
operation of the post filter adsorbers will allow us to assess microbiological effects
associated to COT removals and THM formation potentials. Includes 7 references, tables, figures.