Abstract
The modeling of wastewater treatment processes can be accomplished using multivariate statistical techniques on datasets containing numerous variables that describe the operating status of the processes. The modeling conducted in this study includes the prediction of the overall effluent quality of a wastewater treatment plant (WWTP) with a separated storm sewer, a Wetlands Bioreactor and a Wastewater Lagoon System. The models developed predict BOD, CBOD, TSS, TKN, NH3 and Phosphorus content. Statistical Analysis Software (SAS) was used to develop models using the Generalized Linear Model procedure. The WWTP and Wetlands models used influent quality parameters, atmospheric conditions and flow rate information to predict effluent quality while the Lagoon System required the inclusion of effluent variables for model development. Phosphorus concentration in the WWTP influent exhibited a negative correlation with BOD and CBOD levels in the effluent. Similarly, the effluent phosphorus content was negatively correlated with influent BOD and CBOD levels.