This thesis describes the development of enhanced biological phosphorus removal (EBPR) in a lab-scale sequencing batch reactor (SBR) and the release of phosphorus during the storage and thickening of sludge produced in this reactor. In the first phase of the experimental work a fast start-up method for EBPR development was established by the addition of a pure culture of Acinetobacter lwoffi to a conventional activated sludge. Investigations revealed that the performance EBPR depended on the combination of influent COD and phosphorus values and that in the investigated range, EBPR functioned independently of the sludge retention time. Low dissolved oxygen levels had no effect on the phosphorus removal properties of the sludge. The second phase of the experimental work involved the investigation of the phosphorus released during sludge handling. It was found that phosphorus resolubilisation during sludge treatment took place in three distinct phases which included an initial period of extremely low phosphorus release. Alterations of the reactor influent and operational parameters and the sludge characteristics, affected the amount of phosphorus released during anaerobic storage and gravity thickening. It was found that for short retention times in the sludge processing units (1-48 hours), decreasing the influent phosphorus concentration, increasing the oxidised nitrogen content of the excess sludge and wasting the excess sludge from the aeration tank decreased the amount of phosphorus resolubilised. For longer retention times (2-7 days), it was found that increasing the influent COD, having a lower total phosphorus sludge content, higher sludge "stabilisation" rates and quiescent conditions of storage, decreased the amount of phosphorus released.

This work reports the pilot-scale investigation of various anaerobic reactor systems treating a fruit washing wastewater. An open cell reticulated foam was used as a biomass support media (BSM). The foam pads (25 mm cubes) were randomly packed in the 2.5 m- 3 reactor with an unpacked section beneath the bed. Four general operational regimes were evaluated. These were: single and two stage operation, with and without effluent recycle. Performance was monitored throughout each run in terms of maximum COD loading rate and minimum attainable hydraulic retention time. Biomass concentrations, both within the media and freely suspended between the biomass support particles were measured on samples from each operating regime, their acetoclastic activity being determined in a laboratory test. A method was developed to ascertain whether a difference in biomass activity existed between the outside of an individual biomass support particle and at the centre of the particle, using a radioactively labelled substrate. It was concluded that a two stage system without recycle provided the best performance with respect to the the maximum attainable loading rate (11.6 kgCOD.m- 3 .day). This was approximately twice that for any of the other systems tested. The minimum hydraulic retention time corresponding to this loading was approximately 1.0 d. The superior performance of the two stage system without recycle was attributed to the increased acetoclastic populations brought about by the pre-acidified feed and the plug flow removal kinetics exhibited in reactors without recycle. Two stage systems produced higher levels of biomass in the reactor than their single stage counterparts and a large proportion of the total biomass inventory was present as suspended growth in systems without recycle. Tracer studies showed that the actual HRT was much less than that calculated from flow rate and reactor volume, indicating that large areas of the reactor were not accessible to the substrate. Experiments investigating activity gradients in the BSM indicated that a significant difference existed between the acetoclastic activity of biomass at the centre of a colonized particle and that on the surface. It may be concluded that substrate diffusional limitations played an important role in determining the performance of this type of biomass support. Electron microscope examination of BSP fragments gave little information other than the existence of both attached and suspended growth. Most of the bio mass was present as a dense fibrillar network.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.