The Mathematical Model Of The Volatile Fatty Acids Production From Sewage Sludge By Anaerobic Fermentation

With the rapid increasing amount of sewage sludge, searching for an effective technology for sludge treatment becomes a bottleneck in the field of environmental engineering. Volatile fatty acids (VFAs) production from sewage sludge by anaerobic fermentation is an effective way to realize the harmless treatment and recycle of resource in sludge. In this study, using sewage sludges as substrates, the experimental and mathematical models were investigated under the different key deciding factors. The effects of pH values, sludge concentrations, substrate C/N ratios were explored and their optimal conditions were obtained during the anaerobic fermentation. The appropriate mathematical method for the parameters estimation and the model were developed. Finally, the Anaerobic Digestion Model No.1 (ADM1) established based on the lab-scale experimental results was validated under the condition of large scale demonstrative VFAs production engineering. The main results of this study were shown as follows:(1) The anaerobic fermentation experiments were conducted under different pH conditions by using the thermo-alkaline pretreated liquid of sewage sludge. The dissolution rates of insoluble and soluble proteins and carbohydrates at pH 5.0,7.0 and 10.0 were calculated. By using AQUASIM 2.0 software to fit the experiment data, the dissolution rules of soluble proteins and carbohydrates fitted the Monod equation and that of insoluble proteins and carbohydrates fitted Contois form Based on this ADM1 was modified to describe the hydrolysis process of the organic substrate in the anaerobic fermentation system well and the R2 values are all over 0.99.(2) The effects of pH on the VFAs production by anaerobic fermentation were investigated. With variation of pH from 3.0 to 12.0, the experiment data were measured and simulated. With the increasing of pH values, the VFAs production increased, indicating that the alkaline pH was favorable for the VFAs accumulation. The hydrolysis kinetic rates were estimated and the relationship between the kinetic constants and pH was fitted. The results showed that the linear function estimated the Monod hydrolysis constants under different pH values in sludge anaerobic digestion, which provided a reference value for other studies.(3) The classical particle swarm optimization algorithm (PSO) was improved with the addition of the balance factor K, which was more suitable for estimating biochemical parameters. The improved particle swarm optimization algorithm was applied for estimating the kinetic constants such as Monod constant, hydrolysis rate in ADM1 model. The estimated kinetic constants fitted the simulation results quite well.(4) The effects of different sludge concentrations on VFAs production by anaerobic fermentation were studied. With the increasing sludge concentration from 15g/L to 35g/L, the VFAs increased accordingly. Key kinetic parameters were determined through sensitivity analysis of VFAs generation under the different sludge concentrations, and estimated by the ADM1 model. The modeling values fitted experimental result accurately. The linear equation fitted by MATLAB7.0 described the relationship between concentration and hydrolysis rate.(5) The pretreated sewage sludge liquid was used as substrate for volatile fatty acids (VFAs) production by adjusting the different C/N ratios via anaerobic fermentation. The activities of the key enzymes and the concentrations of acetic acid, propionate and butyrate were measured. With the regression analysis and the function fitting, the function model among C/N ratio, enzyme activities and the acid production was developed. With the R2 (goodness fit) of the model over 0.9, and the less sum of squares, the binary cubic polynomial model was evaluated as the best function model for simulating the relationship of C/N ratio, enzyme activity and VFAs production. The Surface Modeling of key enzyme, VFAs production on C/N ratio conditions was also established. The model can well describe the influence of C/N ratio on the key enzymes and VFAs production in anaerobic fermentation of sludge. The function model provides an effective way to predict the enzyme activity and VFAs fermentation type at different C/N ratios, and is helpful for the future experimental research and the scale up of practical project of VFAs production from sewage sludge.(6) To modeling the free ammonia inhibition in high solid sludge fermentation, the anaerobic digestion model No.1 (ADM1) was modified to simulate the VFAs generation in batch, and semi-continuous sludge. The ADM1 was operated on platform AQUASIM 2.0. Three kinds of inhibition forms, eg. simple inhibition, Monod form and non-inhibition, were integrated into the ADM1 and tested with the real experimental data at batch and semi-continuous fermentation, respectively. The improved particle swarm optimization (PSO) was used for the kinetics parameters estimation by software MATLAB 7.0. In the modified ADM1, the Ks of acetate is 0.025 kgCOD/m3, the km,ac is 12.51 kgCOD/kgCODd, and the K 3 is 0.02, respectively. The results showed that the simple inhibition model can simulate the VFAs generation accurately while the Monod model is the better inhibition kinetics form in semi-continuous fermentation in pH10.0. Finally, the modified ADM1 can successfully describe the VFAs generation and ammonia accumulation in a 30 m3 full scale sludge fermentation reactor, indicating that the developed model can be applicable in high solid sludge anaerobic fermentation.(7) Based on the above results, the modified ADM1 was validated in the scale-up VFAs production engineering from sludge anaerobic digestion. In a demonstrated VFAs production engineering in Wuxi Shuofang wastewater treatment plant with the capacity at 3 m3 (85% water content), the VFAs concentrations, sludge pre-treatment and SCOD release were measured under the conditions of high solid sludge concentration in the start-up and operation stages. The VFAs production, total nitrogen and ammonia concentration were predicted by the modified ADM1 model based on the experimental data. The application of the model predictions indicated the feasibility of large-scale sludge anaerobic fermentation.