Plappally, Anand K

2015-09-25, Gupta, Sandeep, Singh, Rattandeep, Chakraborty, Prodyut Ranjan, Sharma, Rakesh Kumar, Soboyejo, A. B.O., Wei, Xiaohua, Plappally, Anand K

Abstract: Empirical stochastic multi-variable models for prediction of treatment efficiency of wetlands are presented in this article. Wetlands of seven different shapes are visualized using tracer studies. Two different variants of experiments are carried out. Numerous flow rate variations are performed keeping surface area of the wetland constant. The experiment is also carried out with a variation in volume of the wetland which helps to study the effect of flow height on the hydrodynamics within the wetland. A multi-variable model for treatment efficiency in terms of change in tracer concentration as a function of shape, volumetric height of water within the wetland, time, and mass flow rate is considered. Further, another set of experiments is performed studying the treatment efficiency in terms of electro-kinetic parameters. This involves measuring the pH, turbidity, temperature, electrical conductivity, total dissolved salts at inlet and outlet and residence time with varying flow rate, and height of water for the seven different wetland models under study. The electro-kinetic parameters changes due to difference in concentration of the tracer dye which simulates impurities. In this case, treatment efficiency is expressed as a function of the above-discussed electro-kinetic variables, time variation, water height, as well as variation in the mass flow rate. The stochastic multi-parameter models, thus, empirically derived in the above two cases have high coefficient of determination. The models thus derived may be used as a tool for quick analysis of treatment efficiency of any shape and size of a three-dimensional wetland.

2015-09-25, Singh, Rattandeep, Gupta, Sandeep, Raman, S., Chakraborty, Prodyut Ranjan, Sharma, Puneet, Sharma, Rakesh Kumar, Brown, Larry C., Wei, Xiaohua, Plappally, Anand K

Abstract: A numerical visualization study of wetlands is detailed in this article using finite volume methods. The aim of this study is to model treatment efficiency of the wetlands in terms of the residence time distribution function. Shape and depth of wetlands are critically analysed to find the optimal flow requirement for effective treatment. Laminar three-dimensional flow dynamics is used to simulate the slow water flows that occur in treatment wetlands. Slow inlet flows are assumed. Dye is used as the tracer to characterize the hydrodynamics within the wetlands. Three different geometrical configurations, namely square, square with two islands, and triangle, respectively, are simulated. The variation in the tracer concentration is studied as a function of recirculation volumes, flow rates, time and depth of the wetland for each of the wetland shapes. The change in the variation of tracer concentration at inlet and exit helps to assess treatment effectiveness. In another case, glycerine is used to simulate sewage flow. Plug flow is prominent in sewage-laden wetlands. The results obtained from the above-illustrated case studies are compared with each other to assess the reproducibility of the optimal flow model. Multi-parameter regression models for residence time distribution functions are derived to characterize flow through constructed wetlands of different shapes.