Numerical and Experimental Study of Internal Hydraulic Behavior of Horizontal Sub-Surface Flow Constructed Wetland in Different Flow Distributions

Horizontal subsurface flow constructed wetlands have long been applied to improve water or wastewater quality. Previous studies on wetland systems have focused on trying to comprehend the processes leading to the removal of pollutants. Comparatively, there have been fewer studies dedicated to the assessment of flow distribution on hydraulic behavior through the wetland. Researchers declared the aspect ratio (length to width ratio), inlet-outlet configuration, the size of the porous media and the loading rate of constructed wetland could influence hydraulic retention time (HRT). Su et al. (2009) have stated that in free water surface constructed wetlands, when the aspect ratio is greater than 5, the hydraulic efficiency will reach 0.9, or even higher. If the project site or field area cannot meet the theoretical standard, the recommended aspect ratio is higher than 1.88 to ensure some hydraulic efficiency greater than 0.7. The present study was an attempt to analyse, with the aid of 3D numerical simulation and tracer study, how flow distribution affected hydraulic behavior by using 3 different input flow layouts. The treatment system consisted of a horizontal subsurface flow in a constructed wetland having an aspect ratio of 6.5 and the bed slope of one percent. The geometry of this system, which was 4 m wide × 26 m long × 1 m deep, was planted with Phragmites australis. Inlet configurations were selected as a variable parameter. Three different inlet flow configurations including midpoint-midpoint (A), corner- midpoint (B) and uniform-midpoint (C), with the same fixed outlet configurations, were studied. The average flow discharge in each configuration was 6.58, 6.52 and 6.4 m3/day, respectively. Dye tracer was used to draw retention time distribution curves in each configuration for assessing the internal dispersion, short-circuiting and hydraulic parameters such as effective volume rate which is derived by division of mean retention time per nominal retention time. The 3D model presented, which was built on the Comsol Multiphysics platform, was implemented for fluid flow to show internal hydraulic patterns in the system. Hence, the hydraulic model used the Darcy equation to simulate a stationary water flow through the bed. The simulations were verified by using real data obtained from the existing constructed wetland. It was mostly used to show pressure throughout the system for each configuration of the inlet and the outlet. The mean retention time for each configuration was found to be 4.53, 3.24 and 4.65 days, respectively. A marked reduction of the mean hydraulic retention time signified leaving tracer concentration from the outlet rapidly, high short-circuiting and dead volume and finally defective treatment process influenced by changing the inlet to the corner. According to tracer breakthrough curve, the effective volumes for configurations A and C were 87.5%, as compared to 62.1% for the configuration B. The two-day difference of tpeak between configurations 2 and 1, and 3 was probably due to the establishment of preferential streamlines resulting in short-circuiting and areas of dead volume in the system. The value of tpeakis related to dispersion, in the sense that a retention time distribution curve with a small peak time generally contains low dispersion. Simulation results showed the pressure difference from the inlet to the outlet ranged from 12-14, 14-15 and 10-13 cm H2O for A, B and C layouts, respectively. It was shown that the maximum pressure gradient occurred at the outset of the influent wastewater at the inlet, and it was gradually reduced to the lowest values at the outlet ports. Consequence of surface pressure demonstrated uniform pressure from inlet toward outlet at configuration C. Simulated streamlines approved this result, while range of high and low pressure area at configuration B was the most. There was a strong association between tracer experiments and simulation works. One of the major findings of this study was the significant shorter hydraulic mean retention time of the corner inlet setup. There are many that may cause these effects, although short-circuiting may be the primary one. A large low-pressure zone appeared at the opposite corner that was neither inlet nor outlet in this configuration. This paper investigated the hydraulic performance and short-circuiting effects on water flow due to three different inlet patterns (i.e. midpoint, corner, and uniform) in horizontal subsurface flow wetlands based on dye tracer measurements and numerical modeling. The results showed that the uniform inlet could provide the highest hydraulic efficiency (i.e. longest hydraulic retention time, HRT), in comparison to other two setups. The performance of the three different layouts was also investigated in terms of hydraulic parameters. Short-circuiting was influenced by lower hydraulic retention time, leading to inadequate treatment. Uniform-midpoint and midpoint-midpoint yielded the best effective volume as compared to the corner-midpoint. It was demonstrated that these two cases increased dispersion and used the whole capacity of the constructed wetland for the treatment process. The most important result of this paper was the evaluation of internal hydraulic pattern thorough the wetland, something not investigated in previous research works. Based on the simulation results, the spatial pressure distribution in wetland cells was depicted. Finally, it can be concluded that the best configuration of inlet-outlet layout based on both numerical simulations and physical experiments is uniform-midpoint. Meanwhile, midpoint-midpoint is preferable to corner-corner by all performance criteria.