INTEGRATED DESIGN OF SECONDARY CLARIFIER AND BIOLOGICAL NITROGEN REMOVAL PROCESS CONFIGURATION FOR MAXIMUM TREATMENT CAPACITY

The Sha Tin Sewage Treatment Works (STW) is regarded as the largest biological STW in Hong Kong. The existing Stage I/II is currently a 150,000 m3/day 4/5-stage Bardenpho process for biological nitrogen (N) and phosphorus removal. This capacity was much reduced from the original design capacity (conventional process) of 205,000 m3/day, which must be restored to 205,000 m3/day (ADWF) in the near future. A much more volume-efficient Modified Ludzack-Ettinger (MLE) process configuration with bioselector for biological N removal only is selected to meet the more stringent effluent NH4-N standard of < 5 mg/L. Only the existing aeration tank and secondary clarifier hydraulic volumes could be used because of the severe space constraints for the predicted near-term growth in the catchment area.In order to utilise the existing civil structures in a highly efficient, and perhaps innovative, manner, the design and operation of both the MLE process bio-selector and the 30 year-old secondary clarifiers must be integrated. A five-month bench-scale study was conducted to assess the inhibitory effects of the high and variable saline concentration on the biological nitrification-denitrification capacities. The high salinity levels result from seawater toilet flushing (up to 33% of the influent). It was demonstrated that the selected MLE process could meet the required design requirements with a spare 33% extra capacity despite inhibition of up to 20% on the nitrification rates. The effect of the bio-selector and the plug-flow configuration on control of foaming by nocardioforms and low F/M bulking was also verified. However, the existing secondary clarifier with or without modifications for improved hydraulics (centre feed well optimization and Crosby baffle), as indicated from the computer hydraulic modelling, could not meet the design requirements. This led to the reduction of the un-aerated mass fraction of the MLE system from 33% to 25%.A combination of physical and computer modelling was demonstrated to be a valuable aid. The bench-scale (or pilot-scale) physical model can predict the design kinetic parameters and determine the characteristics of the complex microbial community that consumes pollutants, performs nitrification-denitrification, forms flocs for solids-liquid separation, and causes bulking and foaming when overgrowth of the filamentous bacteria occurs. Computer modelling of the biological process and secondary clarifier allowed the performance to be predicted accurately for conditions either not tested or impossible to obtain.Without these integrated considerations, the design and operation may not have been focussed on both the un-aerated configuration of the MLE process and the capacity improvement of the secondary clarifiers. A large un-aerated mass fraction and a conservative estimate of the maximum growth rate of the nitrifiers would have been used, resulting in an unnecessarily long sludge age and high operating MLSS concentration. The secondary clarifier capacity, as indicated by the overflow rate, would have been assumed in order to meet the required capacity. The actual result would have been a nonoptimal system resulting in an unnecessary capital investment and construction duration for larger or more efficient secondary clarifier volumes.