Academic
Publications
OPTIMIZATION OF RESIDENCE TIME IN CONSTRUCTED FREE WATER SURFACE WETLANDS THROUGH BATHYMETRIC DESIGN

OPTIMIZATION OF RESIDENCE TIME IN CONSTRUCTED FREE WATER SURFACE WETLANDS THROUGH BATHYMETRIC DESIGN,R. Michael Conn

OPTIMIZATION OF RESIDENCE TIME IN CONSTRUCTED FREE WATER SURFACE WETLANDS THROUGH BATHYMETRIC DESIGN  
BibTex | RIS | RefWorks Download
Constructed free water surface (FWS) wetlands have become a more popular method of storm water treatment used by practitioners, due to the increasing need for ecological and economically feasible means of detaining and treating urban storm water runoff. Constructed (and natural) wetlands behave similarly to the physiochemical and biological reactors used for treating wastewater, and are thus governed by the same physical laws and parameters as mechanical reactors. A principal parameter used in establishing the treatment effectiveness of reactors is the residence time. The mean residence time, or detention time, for a reactor is the average amount of time the material being treated spends in the reactor. Because the processes that are occurring within a wetland are all a function of time, maximizing the residence time will ultimately improve the treatment. The simplest method of increasing detention time is by increasing the wetland volume. However, space and cost limitations indicate that research focused on finding ways to improve and increase residence time of wetlands is warranted. The objective of this work was to establish relationships between wetland bathymetry and hydraulic detention time using a model that solves the two dimensional hydrodynamic (St. Venant) shallow water flow equations. Simulation results were used to develop design guidelines for maximizing wetland hydraulic residence time, and thus storm water treatment, through specifically designed bottom topography. Design guidelines for the physical features of the wetlands are currently qualitative, and based primarily on aesthetic and ecological purposes (e.g., Schueler, 1992). The research described herein utilizes and builds on the research and experience presented by previous researchers and focuses on maximizing HDT with variations in topography within the storm water wetlands. METHODS Our research was completed in two stages. Preliminary simulations first were made using generated topographies to simulate the effects of various topographic features. Simple rectangular wetlands were simulated in order to focus on the basic topographic features influencing HDT. The results from these simulations were then used to develop design procedures. The developed procedures were used to design a storm water treatment wetland that will be built on the University of Idaho campus in the near future. The numerical model used in this work solves the 2-D hydrodynamic equations using a modified MacCormack, explicit finite difference scheme. The model was originally developed to solve the 2-D hydrodynamic equations for simulating overland flow with microtopography and interactive infiltration (Fiedler, 2000). The model allows for spatial variation in topography, infiltration and surface friction over the domain. However, for the purpose of this research only the effects of topographic features were examined. Uniform flow resistance was assumed, calculated using the Darcy-Weisbach equation. For laminar flow there is a linear relationship between friction factor (f) and Reynold's number Re (Woolhiser, 1975) which is written
Published in 2003.
Cumulative Annual