engleski

PARAMETER SENSITIVITY ANALYSIS OF THE ACTIVATED SLUDGE MODELS FOR WASTEWATER TREATMENT

Sensitivities of activated sludge models parameters were analyzed using global sensitivities analysis according to Fourier Amplitude Sensitivity Test (FAST). Parametric sensitivity analysis of the activated sludge models (ASM1, ASM3 and ASM2d) was performed, in order to define the most important parameters of each model and to performe comparison of each other. ASM models were developed for better control of the wastewater treatment process and modification of biological nutrient removal. The proposed models present world standard for numerical modelling of activated sludge system (Henze et al., 1987, Henze et al. 1995, and Gujer et al., 1999). ASM1 model includes 13 state variables, 8 processes and 19 parameters, ASM3 model includes 13 state variables, 12 processes and 36 parameters. Both models only describe removal of COD and nitrogen which are divided into fractions, presented by state variable. Whereas in ASM1 all state variables are directly influenced by change in a parameter value, in ASM3 that influence is considerably lower, alloving a better identification. ASM2d is a further development of ASM2 with the addition of denitrifying activity of Phosphorus Accumulating Organisms (PAOs) which should allow a better description of phosphate and nitrate dynamic. ASM2d model includes 19 state variables, 19 processes and 43 parameters. The ASM2d classifies biomass into three groups (ordinary heterotrophs, nitrifiers and PAOs). COD components in the model are classified as soluble and particulate. The soluble COD components are either inert or biodegradable. Application of FAST algorithm provides sensitivities to large and simultaneous change of the complete set of the model parameters (Saltelli et al., 2005). Results of the FAST analysis have application in identification of metabolic regulation and potentially elucidate the main effect in enzyme regulation. On the other hand, results of the global sensitivity analysis are important for model improvement and parameter optimization.

wastewater treatment; activated sludge model; global sensitivity analysis; FAST

nije evidentirano