Modeling organic matter and nitrogen removal from domestic wastewater in a pilot-scale vertical subsurface flow constructed wetland

ABSTRACT

Constructed wetlands have become an attractive alternative for wastewater treatment. However, there is not a globally accepted mathematical model to predict their performance. In this study, the VS2DTI software was used to predict the effluent biochemical oxygen demand (BOD) and total nitrogen (TN) in a pilot-scale vertical flow constructed wetland (VFCW) treating domestic wastewater. After a 5-week adaptation period, the pilot system was monitored for another 6 weeks. Experiments were conducted at hydraulic retention times (HRTs) in the range of 2–4 days with Typha latifolia as the vegetation. The raw wastewater concentrations ranged between 144–430 and 122–283 mg L⁻¹ for BOD₅ and TN, respectively. A first-order kinetic model coupled with the advection/dispersion and Richards' equations was proposed to predict the removal rates of BOD₅ and TN from domestic wastewater. Two main physical processes were modeled in this study, porous material water flow and solute transport through the different layers of the VFCW to simulate the constructed wetland (CW) conditions. The model was calibrated based on the BOD₅ and TN degradation constants. The model indicated that most of BOD and TN (88 and 92%, respectively) were removed through biological activity followed by adsorption. It was also observed that the evapotranspiration was seen to have a smaller impact. An additional data series of effluent BOD and TN was used for model validation. The residual analysis of the calibrated model showed a relatively random pattern, indicating a decent fit. Thus, the VS2DTI was found to be a useful tool for CW simulation.

KEYWORDS:

• BOD
• constructed wetland
• nitrogen
• VS2DTI
• wastewater treatment

Acknowledgments

The authors would like to thank the servicing unit of the Faculty of Pharmaceutical Sciences, University of Cartagena, for the parameter analysis. The authors acknowledge the contributions to the project made by Jairo Romero, José Ardila, Edwin Simancas, Juan David Herrera, Delia Curiel, and Roque Palomino.

Funding

The financial support of the University of Cartagena, Colombia, COLCIENCIAS, the Natural Sciences and Engineering Research Council of Canada (NSERC), Ontario Trillium Scholarship (OTS) Program, and Ryerson University is greatly appreciated.