For biological nutrient removal (BNR) systems designed to maximize nitrogen removal, the effluent total nitrogen (TN) concentration may range from 2.0 to 4.0 g N/m3 with about 25–50% in the form of organic nitrogen (ON). In this study, current approaches to modeling organic N conversions (separate processes vs. constant contents of organic fractions) were compared. A new conceptual model of ON conversions was developed and combined with Activated Sludge Model No. 2d (ASM2d). The model addresses a new insight into the processes of ammonification, biomass decay and hydrolysis of particulate and colloidal ON (PON and CON, respectively). Three major ON fractions incorporated are defined as dissolved (DON) (<0.1 µm), CON (0.1–1.2 µm) and PON (41.2 µm). Each major fraction was further divided into two sub-fractions – biodegradable and non-biodegradable. Experimental data were collected during field measurements and lab experiments conducted at the ‘‘Wschod’’ WWTP (570,000 PE) in Gdansk (Poland). The accurate steady-state predictions of DON and CON profiles were possible by varying ammonification and hydrolysis rates under different electron acceptor conditions. With the same model parameter set, the behaviors of both inorganic N forms (NH4-N, NOX-N) and ON forms (DON, CON) in the batch experiments were predicted. The challenges to accurately simulate and predict effluent ON levels from BNR systems are due to analytical methods of direct ON measurement (replacing TKN) and lack of large enough database (in-process measurements, dynamic variations of the ON concentrations) which can be used to determine parameter value ranges.
Keywords: ASM, dissolved organic nitrogen, DON, mechanistic model, nitrogen removal