Multivariate regression analysis of refractive index of activated sludge metabolic status and treatment effectiveness

The metabolic state of activated sludge significantly influences the treatment efficacy of biological wastewater treatment systems. However, conventional microbial activity detection methods are cumbersome and inherently lagging. To address this limitation, a rapid refractive index (RI) detection method was employed to characterize the metabolic state of activated sludge by monitoring surface RI changes on polyvinyl alcohol (PVA) gel beads loaded with activated sludge microorganisms within a completely mixed aeration system. Preliminary continuous monitoring over 100 days in the completely mixed aerated system established good consistency between RI trends and changes in COD_Cr and ammonia nitrogen removal efficiency. Single-factor experiments were conducted by altering operating conditions in the completely mixed aeration system (such as dissolved oxygen, organic loading rate, temperature, pH, salinity, and influent Cu²⁺ concentration) to investigate their respective impacts on treatment performance. Orthogonal experiments were conducted to continuously monitor RI difference parameters (RI maximum value (MaxRI), time to reach maximum value (T_MaxRI), and fluctuation amplitude (ΔRI)) alongside system treatment efficiency. Correlations were analyzed, fitting equations were established, and the RI index was employed to predict treatment performance in completely mixed aeration systems. The results indicated: 1) Variations in RI difference caused by temperature, influent Cu²⁺ concentration, and salinity exhibited more pronounced effects than other factors. 2) Orthogonal experiments revealed that under multi-factor variations, COD_Cr and ammonia nitrogen removal rates could be modeled by the ternary regression equation: Y=B₀+B₁MaxRI+B₂T_MaxRI+B₃ΔRI. The coefficient of determination (R²) was 0.945 7 for COD_Cr removal and 0.613 0 for ammonia nitrogen removal. The RI indicator exhibited strong correlation with treatment efficacy for both COD_Cr and ammonia nitrogen. Thus, this RI indicator enables rapid and accurate prediction of treatment performance in completely mixed aeration systems.