We found anoxic zones in aerated activated sludge flocs, and demonstra
ted denitrification under normal operating conditions. Sulfate reducti
on was not found. Micro-environments and microbial conversions in floc
s from bulking and non-bulking activated sludge were determined with m
icrosensors for H2S, O-2, NO2- and NO3-. Denitrification and sulfate r
eduction rates were measured with N-15- and S-35-tracer techniques. We
showed that under normal reactor conditions (ca. 20% air saturation)
anoxic zones develop within flocs allowing denitrification. The denitr
ification rates amounted to 40% of the rates under anoxic conditions.
At 100% air saturation no anoxic zones were found and no denitrificati
on occurred. However, in flocs from bulking sludge (at 20% air saturat
ion) anoxic zones were absent and denitrification did not occur. In bu
lking sludge only at total anoxia was denitrification found. Confocal
microscopy showed that flocs from bulking sludge were much looser than
those from non-bulking sludge. The absence of anoxic zones and of den
itrification was attributed to the open flee structure, allowing advec
tive oxygen transport. Sulfate reduction was not detected in any of th
e sludges tested by microsensors or by tracer techniques even under an
oxic conditions. This indicates that the sulfur cycle (sulfate reducti
on and sulfide oxidation) does not play a role in mineralization proce
sses and bulking in activated sludge. Preliminary molecular work tin s
itu hybridization with the 16S-rRNA probe SRB385) indicated the presen
ce of small amounts of sulfate reducing bacteria in all sludges. Eithe
r the probe is not specific or the sulfate reducers present are not ac
tive under reactor conditions. (C) 1998 IAWQ. Published by Elsevier Sc
ience Ltd.