Search

A pilot plant of microbubble ozonation and biological treatment was used to treat chemical wastewater in depth, and its long-term operation performance was evaluated and predicted by a neural network model. The results showed that the chemical wastewater could be treated stably and effectively in the pilot plant, and the average removal rates of COD and UV254 in the phase of microbubble ozontion were 35.1% and 52.5%, respectively. The mass ratio of ozone dosage to influent COD amount was an important parameter influencing the operation performance, which should be controlled as 0.5 for better performance. The wastewater biodegradability could be improved effectively by microbubble ozonation, BOD5/COD could increase from 0.10 to 0.33 and the ratio of BOD5 generated amount to COD removed amount could reach to 37.23%. Then biological treatment was applied for the improved water, and the contribution of BOD5 removal to COD removal was more than 80% in biological treatment. Back propagation neural network (BPNN) was used to establish a prediction model of the effect of m on COD removal performance in microbubble ozonation, which showed a stable prediction ability with acceptable accuracy.