APPLYING THE ASM3 MODEL FOR EVALUATION OF CONTROL

MIGUEL MUSSATI,; AGUIRRE, PIO

Rio de Janeiro, Brasil

Congreso; 2nd. Mercosur Congress on Chemical Engineering. 4th. Mercosur Congress on Process System Engineering; 2005

Biological nutrient removal processes for municipal wastewaters have received quite some attention during the past decades, especially due to the introduction of stricter legislation with respect to allowable effluent pollutant discharge limits. The necessity for performance improvement of existing activated sludge wastewater treatment plants (ASWWTPs) has created a need for detailed process models that can be used in the development and evaluation of more efficient activated sludge plant control strategies. The main aim of this work is to compare the plant performance predicted by standard activated sludge models (Activated Sludge Model No 1 (ASM1) and No 3 (ASM3) of the International Water Association (IWA)) when simulating basic control strategies for an ASWWTP case study. The control strategies tested include five basic control loops and seven combined strategies. Compared to the ASM1 model, the ASM3 model predicts in general significantly higher average effluent SNH4 and NTOT concentrations for all basic control loops, assuming for both models the same influent wastewater biodegradable N and COD content, and applying identical controller settings. For strategies with external carbon dosage, the carbon dosage needs predicted by the ASM3 model are generally considerably higher than for the ASM1 model, a point that needs further investigation. When comparing cost performance index (CPI) and effluent quality (EQ), both models often predict similar trends, albeit that the EQ and the CPI for the ASM3 model are generally considerably higher than for the ASM1 model for the same control strategy. Both models predict the lowest CPI for strategy C3 (master DO controller + control of Qrint) but ASM3 does not predict compliance with the effluent ammonium and total nitrogen limits; whereas strategies C6 (master DO controller + carbon dosage) and C* (C6 + control of Qrint) predict the best EQ.NH4 and NTOT concentrations for all basic control loops, assuming for both models the same influent wastewater biodegradable N and COD content, and applying identical controller settings. For strategies with external carbon dosage, the carbon dosage needs predicted by the ASM3 model are generally considerably higher than for the ASM1 model, a point that needs further investigation. When comparing cost performance index (CPI) and effluent quality (EQ), both models often predict similar trends, albeit that the EQ and the CPI for the ASM3 model are generally considerably higher than for the ASM1 model for the same control strategy. Both models predict the lowest CPI for strategy C3 (master DO controller + control of Qrint) but ASM3 does not predict compliance with the effluent ammonium and total nitrogen limits; whereas strategies C6 (master DO controller + carbon dosage) and C* (C6 + control of Qrint) predict the best EQ.int) but ASM3 does not predict compliance with the effluent ammonium and total nitrogen limits; whereas strategies C6 (master DO controller + carbon dosage) and C* (C6 + control of Qrint) predict the best EQ.int) predict the best EQ.