One of the popular domestic sewage treatment process (called step feed oxic-anoxic-oxic process) for nitrogen removal was analyzed in this study by theoretical analysis based on the nitrification and denitrification reaction. Total nitrogen removal efficiency was suggested by considering influent qualities(i.e., ammonia, nitrite, nitrate, alkalinity, and COD). Total nitrogen removal efficiency depends on r (influent allocation ratio). In the case that all influent components are enough, the total nitrogen removal follows equation 100-b/(1+b), when r is 1/(1+b). Finally, it can be concluded that step feed oxic-anoxic-oxic process could be effective for nitrogen removal.

Theoretical total nitrogen removal efficiency and reactor volume ratio in oxic-anoxic-oxic system can be found by influent water quality in this study. The influent water quality items for calculation were ammonia, nitrite, nitrate, alkalinity, and COD which can affect nitrification and denitrification reaction. Total nitrogen removal efficiency depends on influent allocation ratio. The total nitrogen removal follows the equation of 1/(1+b). Optimal reactor volume ratio for maximum TN removal efficiency was expressed by those influent water quality and nitrification/denitrification rate constants. It was possible to expect optimal reactor volume ratio by the calculation with the standard deviation of ±14.2.

Step-feed process for biological nitrogen removal were analyzed numerically for the each unit and final total nitrogen(TN) effluent by water quality management(WQM) model and the results were compared data from these wastewater treatment plants. No bugs and logic error were occurred during simulation work. All of the simulation results tried to two times were obtained and both results were almost same as this model has become good reappearance. It was concluded that most of nitrogen removal occurred in the first oxic tank. Thus the controlling of the first anoxic tank may be more important in term of nitrogen removal. Also each unit of simulation result was kept good relationship with that of measured data. Accordingly this WQM model has good reliance. Finally, WQM model can predict final TN effluent within ±6.0mg/l.