Full scale implementation of biological excess phosphorus removal by means of the activated sludge process commenced in South Africa at the Johannesburg Alexandra Plant in 1974. Originally the plant was designed to operate in the nitrifying extended aeration mode; but by switching off some surface aerators around the common inlet point for the influent and underflow recycle, an anoxic zone was created to induce denitrification. This operational modification worked very successfully. Later additional aerators were switched off to create an anaerobic-anoxic zone at the head of the aeration basin, in this manner attempting to induce the prerequisite conditions for excess biological uptake of phosphorus as put forward by Barnard (i.e. by stimulating P release under anaerobic conditions). This revised operational procedure was not successful due to difficulties experienced in controlling mixing in the anaerobic - anoxic zone. However the experience gained from this investigation together with research findings at laboratory and pilot scale level at the National Institute for Water Research, the University of Cape Town, and Johannesburg instilled sufficient confidence into City Council staff to proceed with the' design of the 150 Ml/d Goudkoppie Plant, initially for nitrogen removal and to subsequently modify it to a 5 stage Phoredox process for both nitrogen and phosphorus removal. Before the plant was commissioned (1976) it was necessary to proceed with extensions to the Northern Works and these were also based on the 5 stage Phoredox process. All the Johannesburg activated sludge plants were comprehensively monitored to determine their performance under cyclic flow and load conditions, The data collected from the Goudkoppie plant were particularly comprehensive and exhaustive. Concomitant with the monitoring program extensive research into the single stage nitrification/denitrification/excess phosphorus removal process was being undertaken by the three agencies mentioned earlier. Of particular interest was the development by the University of Cape Town of a general kinetic model of the single sludge activated sludge process that described carbonaceous degradation, nitrification and denitrification behaviour under cyclic flow and load conditions. This dissertation critically evaluates the predictions of this model against the observed responses on the Goudkoppie and Northern Works.

Includes bibliography.