Integrated Validation Plant: Modelling and optimisation of the biological treatment process

The wastewater infrastructure in Singapore is undergoing a major revamp, which includes the construction of a new water reclamation plant (WRP) in Tuas. This plant aims to maximise the energy recovered from wastewater, with the ultimate aim of achieving energy self-sufficiency. The Integrated Validation Plant (IVP) is a 1000 m3/day pilot plant equipped with an AB configuration. The A-stage biosorption process increases the amount of COD recovered in the primary sludge to increase biogas production, whereas the B-stage membrane bioreactor (MBR) treats the wastewater using activated sludge. The effluent then undergoes reverse osmosis (RO) to produce potable water. This report investigates the biological treatment process in the IVP and aims to answer the following question: Given that most WRP in Singapore uses the Modified Ludzack-Ettinger configuration (MLE; also known as the pre-denitrification configuration) for the B-stage, will a change in configuration reduce net energy consumption while achieving good effluent quality? The research is divided into three main parts with the following objectives: 1. To create and calibrate a Biowin model to simulate the biological treatment process in the IVP 2. To optimise the MBR B-stage by finding the configuration and operating conditions that requires the least influent COD to achieve acceptable effluent quality (in terms of COD, nitrogen and phosphorus), and to determine the appropriate changes in the A-stage (if any) that is required to achieve this effluent quality 3. To explore issues facing the IVP and future developments of wastewater treatment in Singapore – effect of an unaerated deoxygenation tank, feasibility of mainstream Anammox, and modification of the treatment process to reduce chemical dosage in the downstream RO process The Biowin model for the MBR was first calibrated using jar tests and oxygen uptake tests, then validated with profile studies that are being carried out in the MBR regularly. It was found that Biowin is able to model the biological treatment process relatively well, with the major discrepancy being that Biowin is not able to model ammonia adsorption onto activated sludge in the anoxic zone. With the calibrated model, various configurations and operating conditions were simulated to optimise the MBR. It was found that the optimal configuration is the Anoxic-Oxic-Anoxic-Oxic (AOAO) configuration with a step feed of 50% of the influent entering the first anoxic zone and 50% entering the second anoxic zone. The return activated sludge (RAS) flow should be 73% of the influent flow and the DO in the aerated tanks should be maintained at 1 mg/l. As the biosorption tank at the IVP is not functioning well, data was obtained from another biosorption pilot plant in Singapore to calibrate the A-stage. Using this calibrated biosorption model, the optimal MBR configuration, and the typical raw influent characteristics for the IVP, it was determined that a functioning biosorption stage will result in insufficient COD entering the MBR for sufficient phosphorus removal. To achieve the required effluent quality, the best option during steady state operations would be to reduce the sludge retention time (SRT) of the biosorption process from 0.5 days to 0.32 days by increasing sludge wastage as this leads to the highest overall COD recovery. On the other hand, the best method to cope with sudden increases in the nitrogen and phosphorus load in the raw influent would be to partially bypass the biosorption tank and divert some of the raw influent to the primary clarifier as this option results in the fastest response in the effluent quality. As the membrane tanks in the MBR are aerated to scour the membranes and minimise fouling, the mixed liquor in the tanks have a high concentration of dissolved oxygen (DO). A small deoxygenation tank is installed after the membrane tanks to allow DO to deplete before the mixed liquor is recycled back to the anoxic zone at the start of the MBR. However, the deoxygenation tank is not functioning well due to the presence of a cascade overflow from the membrane tanks. Simulations show that the presence of residual DO in the recycle stream leads to a large deterioration the biological phosphorus removal performance of the MBR, mainly due to reduced COD availability for the polyphosphate-accumulating organisms (PAOs). Modifications to reduce the impact of the cascade include the installation of baffle walls, slopes, or overflow pipes. One of the technologies for B-stage that can allow even higher COD recovery is mainstream Anammox for nitrogen removal, which consumes less COD than conventional nitrogen-removal processes. As it works best at temperatures of about 30oC, it may be suitable for implementation in Singapore where the average wastewater temperature is about 31oC. Biowin simulations revealed that mainstream Anammox can occur under the local conditions, with CANON and DEMON configurations being more suitable than the SHARON-Anammox configuration. However, a very high solids removal efficiency in the primary clarifier is required to remove sufficient COD in the A-stage for Anammox to develop in the B-stage. Further studies would be required to determine if this is achievable under local conditions. Lastly, the MBR effluent is sent to a RO process to be reclaimed as potable water. Ammonia and sodium hypochlorite is first added to produce chloramine to prevent biofouling and provide the necessary residual disinfection. it was found that the ammonia concentration in the MBR effluent can be increased sufficiently to remove the need for ammonia dosage. However, the reduced nitrification also resulted in a higher nitrite concentration in the effluent, increasing the amount of sodium hypochlorite needed to produce the required chloramine as sodium hypochlorite also oxidises nitrite into nitrate. In addition, the dosage of anti-scalants can be reduced with better phosphorus removal as the lower effluent phosphorus leads to less scaling potential in the RO process by calcium phosphate.Sanitary EngineeringWater ManagementCivil Engineering and Geoscience