Simulation evaluation of the effects of non-uniform flow and degradation parameter uncertainty on subsurface-flow constructed wetland performance.

Although constructed wetland treatment systems have been used in a variety of applications, uncertainty in adequately determining flow conditions or hydraulic residence times ("hydraulic efficiencies") and degradation model parameters remains a problem with their design. Breakthrough or impulse-type tracer studies in constructed wetlands often result in residence-time distributions exhibiting long skewed "tails" suggesting multiple flow channels or perhaps unrealistically large dispersion factors. A fractional-flow analysis is developed here to quantify possible flow non-uniformity in a subsurface-flow constructed wetland and is then used to assess the effects of non-uniformity and degradation model parameter variability on constituent (for example, chemical oxygen demand) removal. A model application to tracer data developed previously demonstrates how flow non-uniformity alone can account for significant "tailing" and can be related to even moderate estimated dispersion numbers. From the analysis, it is evident that flow non-uniformity is of greater concern than decay parameter uncertainty, and that, from a constructed wetland design and operation perspective, every effort should be made to ensure relative flow uniformity across the constructed wetland.