Treatment of meat processing wastewater for carbon, nitrogen and phosphorus removal in a sequencing batch reactor : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Process & Environmental Technology at Massey University
The typical New Zealand meat processing industry wastewater was treated by a laboratory scale Sequencing Batch Reactor (SBR) to determine an effective operating cycle for biological carbon, nitrogen and phosphorus removal. The Activated Sludge Model No. 1 and Model No. 2 with modifications were used to simulate the treatment of meat processing wastewater using the SBR. The average values of main pollution parameters of the wastewater were characterised as 1390 mg total COD L-1, 755 mg soluble COD L-1, 75 mg L-1 NH3 - N, 145 mg L-1 TKN and 34 mg L-1 TP. The readily biodegradable COD (RBCOD) accounts for 15 - 18% of the total COD, while the inert soluble and particulate portion were 4% each. In order to establish an effective operating cycle for the simultaneous removal of nutrients and organic carbon, different dissolved oxygen (DO) concentrations in the mixed liquor, duration of operating phases and hydraulic retention time (HRT) of a 6 h cycle were tested. The most effective cycle consisted of seven phases. The first two hours of the anaerobic period was followed by the aerobic and anoxic periods. The first aerobic period was maintained at a DO concentration of 0.5 ±0.25 mg L-1 for 1 h, the second aerobic period for 1 h at a DO concentration of 3.75 ±0.25 mg L-1 and the third aerobic period for half an hour at 0.5 ±0.25 mg L-1 DO concentration. A half an hour anoxic period followed the first aerobic period. A settling period of 0.75 h followed the third aerobic period. The last quarter of an hour was for decanting and idling. The solids retention time (SRT) was 15 d, while the HRT was 2.5 d. Greater than 99% removal of biodegradable soluble COD, NH3 - N and PO4 - P was achieved in the effective operating cycle where the TN and TP in the wastewater were reduced to 10 mg L-1 and 1.0 mg L-1, respectively. In addition the soluble COD was reduced to 98 mg L-1. The key kinetic and stoichiometric parameters for ASM 1 and ASM 2 models were determined using batch tests. The heterotrophic maximum specific growth rate, yield coefficient and the half saturation constant were 2.0 d-1, 0.63 mg cell COD (mg COD)-1 and 8 mg L-1 respectively. The maximum specific growth rate of autotrophs was 0.65 - 0.80 d-1. The anaerobic phosphorus removal stoichiometric coefficients were also determined in batch tests. During the anaerobic period, when 1 g of acetate COD was initially present, 1.48 g of PHA COD was stored while 0.48 g of P was released. The batch trials conducted using acetate to assess the influence of Mg2+ in P uptake showed that the Mg2+ could limit the P uptake and the uptake rate could be represented by Monod type kinetics. In the Monod kinetic expression the Mg2+ half saturation constant was found to be 4.7 mg L-1 The molar ratio of Mg2+ with P was 0.21 during the anaerobic period, and 0.33 during the aerobic period. The SBR performance was modelled using ASM 1 and ASM 2 models after the addition of more processes in these models. Ammonification of the soluble organic N process rate was modified in the ASM 1 model. Similarly it was necessary to add anoxic P uptake and anoxic growth processes involving PHA of Bio-P bacteria in the ASM 2 model. Glycogen storage and glycogen lysis processes of Bio-P bacteria were added in the ASM 2 model to understand the involvement of glycogen in P removal. Also a modification was performed to the storage process of poly-P in the ASM 2 model to account for potential Mg2+ limitation in meat processing wastewater treatment for P removal. During the settling period anoxic hydrolysis was assumed to be negligible. The calibrated ASM 1 and ASM 2 models in general well simulated the effluent NH3 - N, NO3 - N and PO4- P of SBR cycles carried out in distinctly different periods of time and in different batch tests. As the calibrated modified ASM 2 model was able to predict the performance of an SBR cycle conducted over a time period of three months, it was used to identify the most promising treatment strategies of the SBR performance. Variation in duration of feed cycle during the first non-aerated mixed period did not affect the effluent NO3 - N, NH3 - N and PO4 - P concentrations significantly. DO concentration of 3.75 mg L-1 during the third aerobic period instead of 0.5 mg L-1 increased the effluent NO3 - N and PO4 - P concentrations. The simulations confirmed that the operating conditions identified in a 6-h cycle period for the simultaneous organic carbon and nutrient removal are effective.