Operational Performance And Mathematical Model Of LMBR Treating Pharmacy Wastewater Under Different Ecological Conditions

Cephalosporin intermediate is the important factor of influencing the effluent COD to meet the discharge standard which is one of the most representative characteristic pollutants in the cephalosporin pharmaceutical wastewater. So it is very necessary to treat the cephalosporin intermediate wastewater effectively. In this paper, we used conventional membrane bioreactor (CMBR) and loop multi-loop membrane bioreactor (LMBR1 and LMBR2) on treatment of the cephalosporin intermediate wastewater. And studied their treatment effect of the pollutants (COD, NH4+-N, TN, Cephalosporin intermediates) in Cephalosporin intermediate wastewater, researched the influence of LMBR treatment effect in different environmental factors,thus it can took as an reference and basic for the actual project.The optimal operation condition and HRT was determined by the performance of 3 different set to treat the cephalosporin intermediate wastewater treatment. The results showed that:the COD, ammonia nitrogen, TN and cephalosporins intermediates removal rate of LMBR2 was significantly higher than the LMBR1 and CMBR with the same HRT. HRT=14h, the average COD ramoval rate of CMBR, LMBR1 and LMBR2 were 88.44%,90.5% and 90.78%, respectivily; the average ammonia ramoval rate of the CMBR, LMBR1 and LMBR2 were 82.9%,85.51% and 85.79%, respectivily; the average TN ramoval rate of the CMBR, LMBR1 and LMBR2 were 60.0%,67.94 and 68.9%, respectivily; the average 7-ACA ramoval rate of the CMBR, LMBR1 and LMBR2 were 76.68%,80.32% and 80.82%, respectivily; However, the performance of pollutants removal will reduce with the HRT reduction.In this study, we also researched the influence of LMBR treatment effect in different environmental factors, and the optimal process parameters were obtained by the experiments. The response surface methodology(RSM) was combined in this research to ascertained the optimal HRT was 14h, the dissolved oxygen is 2.0mg/L and the optimal loading rate of 7-ACA is 0.023 kg7-ACA/m3·d.The models which could simulate and pridicct the performance of LMBR during the stable operation under different ecological factors were constract based on ASM1, multiple linear regression and BP neural network. These models revealing the quantitative relationship between the different environment ecological factors and COD, ammonia nitrogen and 7-ACA removal rate. The three different models were employed to predict the effluent of LMBR and compared the performance of the three different models found that BPNN has the highest accuracy. ASM1, as a mechanistic model, not has a superior performance than the BPNN model, which proves that the neural network has good prospects on the sewage treatment system.