The treatment process at Hammarsdale Waste-water Works has been heavily influenced by the nature of the influent, which is largely of industrial origin, as well as by certain design features of the extended aeration 5-stage Bardenpho process present. In particular, dissolved phosphate, chemical oxygen demand (COD) and colour removal have proved problematic. Limitations of the activated sludge process have been seen as the underlying origin of these problems. Change from a diffused to surface aeration system in 1987 to 1990 conferred greater stability on the biological process but did not prove adequate to produce enhanced phosphorus removal to the extent that the special phosphate standard (1 mg P/l as dissolved orthophosphate) could be met. This paper examines the benefits at full scale of simultaneous precipitation using aluminium sulphate (alum) dosed into either the anaerobic zone or return activated sludge. Apart from producing consistent phosphorus removal to below 1 mg P/l as dissolved orthophosphate, alum dosing at 150 mg/l (relative to raw influent) has given significantly improved sludge settleability. COD and colour removal have not improved significantly (contrary to laboratory-scale tests), although the effect of sludge age in this respect was not investigated. Increased alum dosing (200 mg/l) failed to produce statistically lower effluent COD or colour in a full-scale trial. This study highlighted the need for a more reliable sludge wasting/dewatering practice in order to improve control of sludge age. Theoretical predictions of sludge production were compared to Works data. The unbiodegradable particulate fraction of the influent COD proved to be a critical unknown especially in view of the likely contribution of textile and chicken abattoir waste from local industries. Limited full-scale data suggest that the increase in sludge production due to alum is less than the 30% projected from laboratory trials.
