The aim of this study was to establish a mathematical model for simulating a biological nutrient removal (BNR) process for treating wastewater with different COD:N ratios. Through phylogenetic analysis, the microbial communities in a cyclic activated sludge system (CASS) reactor under different COD: -N ratios were quantitatively characterised, while the necessity of incorporating the microbial interaction mechanism in the extended ASM1 (E-ASM1) was validated. Since the newly introduced parameters affected the original kinetics in the activated sludge model No. 1 (ASM1), the poorly defined parameters were altered based on the batch experimental data taken from a lab-scale CASS and the results of a root-mean-square sensitivity analysis. The simulation results indicated that effluent COD and biopolymer concentrations decreased with an increase in influent COD: -N ratio, while the trend was opposite for nitrification. Compared to the ASM1-SMP-EPS model, the extended ASM1 (E-ASM1) proved to be a preferable reference for the simulations of the CASS performance in substrate degradation and biopolymers formation mechanism, strongly suggesting that the interactive relationship among different bacteria and the biopolymer dynamics should be simultaneously included in this model.