Wetland Module Research Articles

Modelling the recovery time from peak loads in a full-scale horizontal flow wetland in Sicily

The aim of this study was to simulate the water flow and reactive transport of pollutants in a horizontal flow (HF) wetland to better understand the recovery time of the treatment performance for peak load events. For the simulation, the processes-based model HYDRUS and its Wetland Module is used. The system under investigation is the first stage of the 9-years old hybrid treatment wetland of a large retail store, located in Catania, Italy. For the calibration of the hydraulic model, the data of a tracer test was used. The data set of the systems is available for a seven year period including organic matter and ammonia nitrogen. The data was split into a standard event representing low loading conditions and determined peak load events with high loadings. The results show that the response time of the model correlates with the hydraulic retention time from the tracer experiment and indicates that higher peak load concentrations at the inlet of the system lead to a longer recovery time of the wetland.

Hydrogeochemical analysis of processes affecting HCH removal using ZVI-based treatment technology

In the 1960s, ca. 3000–5000 tons of HCH residue were disposed of in an overburden dump at the Hájek kaolin/uranium mine site (Czech Republic). The dump leachate, which has a 136 μg/l average total content of hexachlorocyclohexane (HCH) isomers, discharges into a local creek and contaminates the ecosystem. A full-scale prototype for the treatment of dump leachate comprised three sequential stages: 1) permeable reactive modules filled with macro-zerovalent iron (mZVI), 2) a biosorption module and 3) an aerobic wetland module. After commissioning, HCH removal efficiency reached 95%, but decreased to 70% over the following 230 days, primarily due to the efficiency of the mZVI modules dropping from 76% to 39%. Hydrogeochemical analysis and geochemical modelling revealed that the reduction in efficiency was mainly caused by passivation of mZVI surfaces and clogging of mZVI pore spaces through precipitation of goethite, calcite and rhodochrosite, or siderite and magnetite instead of goethite where boundary conditions for O2 differed. XRD analysis of the solid phase from the mZVI modules confirmed the geochemical modelling results. The major part of the precipitated products (46–66 wt%) comprised a ferric hydroxide amorphous phase, especially in the initial mZVI module inlet, with goethite the second most abundant precipitate (10–40 wt%). Siderite and calcite also mainly precipitated in the initial module inlet. In conclusion, mZVI appears to be a suitable reductant for HCH; however, the longevity of the ZVI-based treatment system was negatively affected by precipitates in the presence of high concentrations of iron, manganese and carbonate species.

Improvement of Soil Water Assessment Tool (SWAT) wetland module for modelling of ephemeral pond hydrology

AbstractEphemeral ponds (EPs) are seasonally flooded isolated wetlands that provide a variety of hydroecological benefits, including the provision of breeding habitat for several amphibian and invertebrate species. However, the lack of their explicit representation in hydrological models limits a comprehensive understanding of their interaction with surrounding landscapes and their vulnerability in the context of human interventions and climate change. The purpose of this research was to improve the isolated wetland module of the Soil Water Assessment Tool (SWAT) to better represent EP hydrology. The changes include (1) representation of groundwater and hypodermic flow as the only inflows from the pond drainage surface, due to the intermittent and negligible presence of inflow from surface runoff in forested ponds, (2) revision of how evapotranspiration within EPs is represented and (3) implementation of distinct volume‐area‐depth relationships for ponds based on their geometrical shape. The accuracy of these improvements was assessed against that of a previous isolated wetland formulation in replicating water depth observations of 10 EPs of a portion of the Kenauk forest (68 km2) in the Canadian Shield of the Outaouais region (Québec, Canada). The comparison results show that the revised SWAT model presented here significantly improves the distinct filling and drying water cycle of EPs (average root mean square error of 0.1 m of the revised model vs. 0.23 m for the original model). Besides, the new module allowed to identify that hypodermic flow, evapotranspiration and seepage to the underlying soil are the main EP source and sinks. The new module also allowed to explicitly quantify the differences in filling/drying pattern of the EPs of the Kenauk forest and unlike the original model structure, the new module was able to closely replicate the interannual variation of spring and annual hydroperiod duration.

Evaluation of AnnAGNPS Wetland Nitrogen Removal Component on Farmed Prairie Potholes

Highlights AnnAGNPS offers a wetland module (AgWET) to simulate farmed prairie potholes soluble nitrogen dynamics. Three key parameters of the AgWET nitrogen removal components show high and very high Relative Sensitivity (Sr ). AnnAGNPS tends to underestimate the soluble nitrogen concentration at the beginning of the cropping season. Abstract. Farmed prairie potholes are common in northern Iowa and have been considered an impediment to agricultural development. Potholes have distinctive wetland characteristics where they might contain a high concentration of soluble nitrogen coming from agricultural upland runoff. With sub-surface drainage underneath them, providing short circuit nutrient transport downstream, Iowa poses a significant threat to nitrogen export to the Gulf of Mexico. Watershed models, such as AnnAGNPS, developed a wetland module allowing users to find the possible BMPs to accomplish nutrient reduction targets. Previous research indicates that AnnAGNPS may be used to mimic inundation depth fluctuations in potholes with moderate success. However, there has been little research on the wetland nitrogen removal components of the model. This study aims to evaluate the AnnAGNPS wetland module's performance in simulating the soluble nitrogen concentration in three farmed prairie potholes by comparing the simulation output with the observed data. The sensitivity analysis of the three key parameters of the model shows high and very high levels of sensitivity, with the temperature coefficient's relative sensitivity ranging from -6.7 to 8.6. Attempts in parameterization on the three targeted parameters indicated that the model poorly simulated the soluble nitrogen concentration. The comparison between observed and simulated data shows that the model fails to predict the high concentration of soluble nitrogen in the early cropping season. This conclusion is consistent with earlier research indicating that AnnAGNPS tends to underestimate the nitrogen concentration during the cropping season. Keywords: AnnAGNPS wetland module, Farmed prairie pothole, Soluble nitrogen concentration.

Modelling the terrestrial nitrogen and phosphorus cycle in the UVic ESCM

Abstract. Nitrogen (N) and phosphorus (P) biogeochemical dynamics are crucial for the regulation of the terrestrial carbon cycle. In Earth system models (ESMs) the implementation of nutrient limitations has been shown to improve the carbon cycle feedback representation and, hence, the fidelity of the response of land to simulated atmospheric CO2 rise. Here we aimed to implement a terrestrial N and P cycle in an Earth system model of intermediate complexity to improve projections of future CO2 fertilization feedbacks. The N cycle is an improved version of the Wania et al. (2012) N module, with enforcement of N mass conservation and the merger with a deep land-surface and wetland module that allows for the estimation of N2O and NO fluxes. The N cycle module estimates fluxes from three organic (litter, soil organic matter and vegetation) and two inorganic (NH4+ and NO3-) pools and accounts for inputs from biological N fixation and N deposition. The P cycle module contains the same organic pools with one inorganic P pool; it estimates influx of P from rock weathering and losses from leaching and occlusion. Two historical simulations are carried out for the different nutrient limitation setups of the model: carbon and nitrogen (CN), as well as carbon, nitrogen and phosphorus (CNP), with a baseline carbon-only simulation. The improved N cycle module now conserves mass, and the added fluxes (NO and N2O), along with the N and P pools, are within the range of other studies and literature. For the years 2001–2015 the nutrient limitation resulted in a reduction of gross primary productivity (GPP) from the carbon-only value of 143 to 130 Pg C yr−1 in the CN version and 127 Pg C yr−1 in the CNP version. This implies that the model efficiently represents a nutrient limitation over the CO2 fertilization effect. CNP simulation resulted in a reduction of 11 % of the mean GPP and a reduction of 23 % of the vegetation biomass compared to the baseline C simulation. These results are in better agreement with observations, particularly in tropical regions where P limitation is known to be important. In summary, the implementation of the N and P cycle has successfully enforced a nutrient limitation in the terrestrial system, which has now reduced the primary productivity and the capacity of land to take up atmospheric carbon, better matching observations.

Capability of HYDRUS wetland module to simulate flow and nitrogen removal processes in pilot-scale treatment peatlands under frost and no-frost conditions

The HYDRUS wetland module is widely used together with the biokinetic model CWM1 to simulate reactive transport of contaminants in constructed wetlands. However, this approach has not been used previously to simulate processes in peat-based wetlands operating in cold climates and treating mining-influenced water. In this study, the goal was to clarify changes in flow, transport, and nitrogen removal processes in cold climate treatment peatlands by assessing the performance of HYDRUS-CWM1. Flow and non-reactive transport of tracer, and reactive transport of ammonium, nitrite, and nitrate, in two pilot wetlands operated under controlled conditions representing frozen (winter) and frost-free (summer) periods were simulated. Model simulation outputs were compared against data obtained from the pilot wetlands and from a full-scale treatment peatland treating mining-influenced water in an Arctic region. Initial peaks in tracer concentration were simulated satisfactorily, but transformation and transport of nitrogen species in treatment peatlands, especially under partially frozen conditions, were modeled with only limited success. Limitations of the model and the assumptions made for the simulations have been discussed to highlight the challenges in modeling of treatment peatlands.

Groundwater-surface water interactions at wetland interface: Advancement in catchment system modeling

Wetlands strongly interact with groundwater and surface water, influencing catchment hydrology and altering water quality. Meanwhile, catchment-scale models are typically unable to simulate groundwater-wetland interactions despite the fact that quantifying groundwater-wetland interactions can assist in better identifying locations for wetlands restoration/creation. We modified an integrated groundwater-surface water model (SWAT-MODFLOW-RT3D) to simulate water and nutrient exchange at the wetland interface. Several modifications were applied to the SWAT wetland module, and a spatial linkage was established between the module and the MODFLOW Drain Package to provide bidirectional water and nutrient exchange between groundwater and wetlands. We applied the modified model (SMR-W) to a tropical catchment in northeast Australia and quantified water and nitrate exchange between wetland, groundwater, and surface water for 28 potential locations. This study demonstrates that when groundwater heads are perched above wetlands bed, significant nitrate discharge from aquifers to wetlands occurs, which should be considered during wetland restoration planning.

Role of macrophyte species in constructed wetland-microbial fuel cell for simultaneous wastewater treatment and bioenergy generation

Abstract The aim of this study was to determine the role of plants on pollutant removal and bioelectricity production in the treatment of municipal wastewater with constructed wetland matrix incorporating microbial fuel cells (CW-MFCs) compared to traditional CWs. Multi-anode unplanted and planted CW-MFCs (Iris pseudacorus, Hyacinth pink, and Phragmites australis) were established in fed-batch mode. CW-MFC modules with established vegetation had high treatment efficiencies with COD, NO3−-N, NH4+-N, and PO43−-P removal of 46.9–51.6%, 94.8–97.4%, 43.2–71.5%, and 96.0–97.6%, respectively, compared to 36.6%, 89.9%, 43.0%, 97.1% in the unplanted wetland module, respectively. More efficient pollutant removal and higher power production were correlated with higher plant growth. The highest maximum power density achieved was 25.14 mW/m2 in the multi-anode CW-MFC planted with Iris pseudacorus at highest plant height (1635 cm). The CW-MFC planted with Iris pseudacorus enhanced NH4+-N removal by 66.2% and significantly contributed to bioelectricity generation by 97.5% compared to the unplanted CW-MFC. The results highlight the significant role of growing plants in the CW-MFC matrix in strengthening the bioenergy output compared with enhancement in wastewater treatment in CW-MFCs. Future studies should focus on improvement in cathode potential over a long-term operation and minimize the negative role of withered plants on the planted CW-MFCs.

Aeration intensity simulation in a saturated vertical up-flow constructed wetland

Simulation and performance results of a saturated vertical up-flow constructed wetland (SVU CW) operated under different operational conditions are presented. The SVU CW consists of two different systems planted with Cyperus alternifolius and Iris pseudacorus, and each system consists of three SVU beds operated in series. The SVU CW operates in continuous aeration (CA) mode using different air-water ratios from 0.5:1 to 4:1. The aerated SVU CW achieves a high (more than 85%) removal of chemical oxygen demand (COD), ammonium (NH4+-N), total nitrogen (TN) and total phosphorus (TP). Furthermore, we simulate the SVU CW using the HYDRUS Wetland Module using the CWM1 biokinetic model under CA mode. According to the simulation results, aeration intensity controls the substrate distribution and growth of bacteria with depth in the SVU CW. Organic matter (OM) and nitrogen are removed in the top region (0–30 cm) of the SVU CW. The root mean square error for COD and NH4+-N is >1.5, whereas R2 is >0.99. A good match between observed and simulated data suggests that the CWM1 model is a suitable tool for simulating various processes and bacterial dynamics in aerated SVU CWs.

Influence of design parameters on the treatment performance of VF wetlands - a simulation study.

The main approach for designing vertical flow (VF) treatment wetlands is based on areal requirements ranging from 2 to 4 m2 per person equivalent (PE). Other design parameters are the granularity of the filter material, filter depth, hydraulic and organic loading rates, loading intervals, amount of single doses as well as the number of openings in the distribution pipes. The influence of these parameters is investigated by running simulations using the HYDRUS Wetland Module for three VF wetlands with different granularity of the filter material (0.06-4 mm, 1-4 mm, and 4-8 mm, respectively). For each VF wetland, simulations are carried out at different temperatures for different organic loading rates, loading intervals and number of distribution points. Using coarser filter material results in reduced removal of pollutants and higher effluent concentrations if VF wetlands are operated under the same conditions. However, the treatment efficiency can be increased by applying more loadings and/or a higher density of the distribution network. For finer filter material, longer loading intervals are suggested to guarantee sufficient aeration of the VF filter between successive loadings.

A coupled surface water storage and subsurface water dynamics model in SWAT for characterizing hydroperiod of geographically isolated wetlands

Wetlands play an important role in watershed hydrology and biogeochemistry. A geographically isolated wetland (GIW) module for Soil and Water Assessment Tool (SWAT) was developed to couple surface water storage and subsurface water dynamics to characterize hydroperiod. The new GIW module includes the following features: (1) a flexible geometric formula to characterize wetland surface water area, volume, and depth; (2) a revised algorithm to account for evaporation from both water and soil surfaces in wetlands; (3) application of the Richards equation to couple surface water storage and subsurface water dynamics; (4) use of Darcy's law with an effective hydraulic conductivity parameter to simulate groundwater discharge. We tested the GIW module using observed daily water level data from four wetlands at two sites (including restored and natural wetlands with and without a low-permeability soil layer) in the Delmarva Peninsula, USA. The results show that the wetland module reasonably reproduced observed water levels for both restored and natural wetlands with and without a low-permeability soil layer. The module was also able to reasonably simulate saturated and unsaturated portions of the soil corresponding to wet and dry periods. The ability of the GIW module to describe inundation conditions for wetlands holds promise to enhance the understanding and quantification of hydrological and biogeochemical roles of GIWs in a watershed context.

Numerical simulation of vertical flow wetlands with special emphasis on treatment performance during winter.

In Austria, single-stage vertical flow (VF) wetlands with intermittent loading are a state-of-the-art technology for treating domestic wastewater. They are designed according to the Austrian design standard with a specific surface area of 4 m2 per person (i.e. 20 g COD/(m2·d)) and thus demand a bigger amount of land to treat the same amount of wastewater compared to intensified technical treatment systems. In order to reduce the amount of land needed, a modified design for VF wetlands has been proposed. The modified design has a specific surface area of 2.5 m2 per person (i.e. 32 g COD/(m2·d)) and it has been shown to be able to meet the Austrian effluent requirements. To allow higher organic loading, more loadings per day but lower volume of a single loading, a constant loading interval, and increased number of openings per m2 are applied. A simulation study using the HYDRUS Wetland Module was carried out to compare the treatment efficiencies of single-stage VF wetlands with classical and modified design. Data from a classical Austrian single-stage VF wetland was used for calibration of the model using the standard parameter set for the CW2D biokinetic model. The influent COD fractionation was calibrated to adapt to the wastewater. The simulations showed a good performance of the modified design compared to a classical VF wetland for COD removal with COD effluent concentrations in winter (effluent water temperature of 4.5 °C) of 35 and 29 mg/L, respectively. The simulation study showed that during high-loading events the VF wetland with modified design has lower maximum NH4-N effluent concentrations. Single-stage VF wetlands with modified design seem to be very effective and allow application of higher organic loads compared to single-stage VF wetlands with classical design.

Simulating vertical flow wetlands using filter media with different grain sizes with the HYDRUS Wetland Module

Abstract In this study, the simulation results of four vertical flow wetland systems using the HYDRUS Wetland Module are presented. The four wetland systems comprise three single-stage pilot scale systems and one full-scale two-stage system. The main difference between these systems is the filter media used, referred to as fine media, i.e., sand with a grain size distribution 0.063–4 mm, or coarse media, i.e., sand with grain size distributions between 1–4 mm, respectively. The water-flow simulation of each system is carried out using the single porosity van Genuchten-Mualem model. A good match between measured and simulated volumetric effluent flow rates could be achieved for all wetland systems. For reactive transport simulations, the CW2D biokinetic model was applied. First, simulations were run using the standard CW2D parameter set. For some systems, adjustments of the parameter set were needed in order to avoid unlimited bacteria growth. To better fit measured COD, NH4-N, and NO3-N effluent concentrations, adjustments of few parameters of the standard parameter set were required. The results show that for the VF wetlands with fine sand, no adjustments of the CW2D standard parameter set were needed, while for systems with coarser filter media as the main layer, the standard parameter set had to be adjusted to match simulated and measured effluent concentrations.

Effect of vegetation type on treatment performance and bioelectric production of constructed wetland modules combined with microbial fuel cell (CW-MFC) treating synthetic wastewater.

An operation of microcosm-constructed wetland modules combined with microbial fuel cell device (CW-MFC) was assessed for wastewater treatment and bioelectric generation. One of the crucial aims of the present experiment is also to determine effect of vegetation on wastewater treatment process and bioelectric production in wetland matrix with microbial fuel cell. Accordingly, CW-MFC modules with vegetation had higher treatment efficiency compared to unplanted wetland module, and average COD, NH4+, and TP removal efficiency in vegetated wetland modules were ranged from 85 to 88%, 95 to 97%, and 95 to 97%, respectively. However, the highest NO3- removal (63%) was achieved by unplanted control module during the experiment period. The maximum average output voltage, power density, and Coulombic efficiency were obtained in wetland module vegetated with Typha angustifolia for 1.01 ± 0.14V, 7.47 ± 13.7mWatt/m2, and 8.28 ± 10.4%, respectively. The results suggest that the presence of Typha angustifolia vegetation in the CW-MFC matrix provides the benefits for treatment efficiency and bioelectric production; thus, it increases microbial activities which are responsible for biodegradation of organic compounds and catalyzed to electron flow from anode to cathode. Consequently, we suggest that engineers can use vegetated wetland matrix with Typha angustifolia in CW-MFC module in order to maximize treatment efficiency and bioelectric production.

Understanding the hydrologic control of N cycle: Effect of water filled pore space on heterotrophic nitrification, denitrification and dissimilatory nitrate reduction to ammonium mechanisms in unsaturated soils

Irrigation practice will be effective if it supplies optimal water and nutrients to crops and act as a filter for contaminants leaching to ground water. There is always a scope for improving the fertilizer use efficiency and scheduling of wastewater irrigation if the fate and transport of nutrients particularly nitrogenous compounds in the soil are well understood. In the present study, nitrogen transport experiments for two different agricultural soils are performed under varying saturation 33, 57, 78% water filled pore space for sandy soil 1 and 52, 81 and 96% for loam soil 2. A HYDRUS 2D model with constructed wetland (CW2D) module could simulate aerobic nitrification and anoxic denitrification well for both soils and estimated the reaction kinetics. A hot spot of Dissimilatory Nitrate Reduction to Ammonium (DNRA) pathway has been observed at 81% moisture content for a loamy sand soil. The presence of high organic content and reductive soil environment (5.53 C/NO3− ratio; ORP=−125mV) results in ammonium accumulation of 16.85mg in the soil. The overall observation from this study is nitrification occurs in a wide range of saturations 33–78% with highest at 57% whereas denitrification is significant at higher water saturations 57–78% for sandy soil texture. For a loamy sand soil, denitrification is dominant at 96% saturation with least nitrification at all saturation studies. The greatest nitrogen losses (>90%) was observed for soil 2 while 30–70% for soil1. The slow dispersive subsurface transport with varying oxygen dynamics enhanced nitrogen losses from soil2 due to lesser soil permeability. This in turn, prevents NO3− leaching and groundwater contamination. This type of modeling study should be used before planning field experiments for designing optimal irrigation and fertigation schedules.

Applying Process-Based Models for Subsurface Flow Treatment Wetlands: Recent Developments and Challenges

To date, only few process-based models for subsurface flow treatment wetlands have been developed. For modelling a treatment wetland, these models have to comprise a number of sub-models to describe water flow, pollutant transport, pollutant transformation and degradation, effects of wetland plants, and transport and deposition of suspended particulate matter. The two most advanced models are the HYDRUS Wetland Module and BIO-PORE. These two models are briefly described. This paper shows typical simulation results for vertical flow wetlands and discusses experiences and challenges using process-based wetland models in relation to the sub-models describing the most important wetland processes. It can be demonstrated that existing simulation tools can be applied for simulating processes in treatment wetlands. Most important for achieving a good match between measured and simulated pollutant concentrations is a good calibration of the water flow and transport models. Only after these calibrations have been made and the effect of the influent fractionation on simulation results has been considered, should changing the parameters of the biokinetic models be taken into account. Modelling the effects of wetland plants is possible and has to be considered when important. Up to now, models describing clogging are the least established models among the sub-models required for a complete wetland model and thus further development and research is required.

Using numerical simulation of a one stage vertical flow wetland to optimize the depth of a zeolite layer.

This simulation study investigates the treatment performance of a compact French vertical flow wetland using a zeolite layer in order to increase ammonium nitrogen removal. For the modelling exercise, the biokinetic model CW2D of the HYDRUS Wetland Module is used. The calibrated model is able to predict the effect of different depths of the zeolite layer on ammonium nitrogen removal in order to optimize the design of the system. For the model calibration, the hydraulic effluent flow rates as well as influent and effluent concentrations of chemical oxygen demand (COD) and NH4-N have been measured. To model the adsorption capacity of zeolite, Freundlich isotherms have been used. The results present the simulated treatment performance with three different depths of the zeolite layer, 10 cm (default), 15 cm and 20 cm, respectively. The increase of the zeolite layer leads to a significant decrease of the simulated NH4-N effluent concentration.

An enhanced SWAT wetland module to quantify hydraulic interactions between riparian depressional wetlands, rivers and aquifers

This study develops a modified version of the Soil and Water Assessment Tool (SWAT) designed to better represent riparian depressional wetlands (SWATrw). It replaces existing unidirectional hydrological interactions between a wetland and a river/aquifer with a more robust bidirectional approach based on hydraulic principles. SWATrw incorporates a more flexible wetland morphometric formula and a connecting channel concept to model wetland-river interactions. SWAT and SWATrw were tested for the Barak-Kushiyara River Basin (Bangladesh and India). Although the two models showed small differences in simulated stream flow, SWATrw outperformed SWAT in reproducing river stages and the pre-monsoon river-spills into riparian wetlands. SWATrw showed that the observed presence of dry season water in the wetland was due to reduced seepage to the local groundwater table whilst continuous seepage simulated by SWAT resulted in the wetland drying out completely. The new model therefore more closely simulates the hydrological interactions between wetlands, rivers and groundwater.

Integrating isolated and riparian wetland modules in the PHYSITEL/HYDROTEL modelling platform: model performance and diagnosis

Mathematical modelling is a well-accepted framework to evaluate the effects of wetlands on stream flow and watershed hydrology in general. Although the integration of wetland modules into a distributed hydrological model represents a cost-effective way to make this assessment, the added value brought by landscape-specific modules to a model's ability to replicate basic hydrograph characteristics remains unclear. The objectives of this paper were the following: (i) to present the adaptation of PHYSITEL (a geographic information system) to parameterize isolated and riparian wetlands; (ii) to describe the integration of specific isolated wetland and riparian wetland modules into HYDROTEL, a distributed hydrological model; and (iii) to evaluate the performance of the updated modelling platform with respect to the capacity of replicating various hydrograph characteristics. To achieve this, two sets of simulations were performed (with and without wetland modules), and the added value was assessed at three river segments of the Becancour River watershed, Quebec, Canada, using six general goodness-of-fit indicators and 14 water flow criteria. A sensitivity analysis of the wetland module parameters was performed to characterize their impact on stream flows of the modelled watershed. Results of this study indicate the following: (i) integration of specific wetland modules can slightly increase the capacity of HYDROTEL to replicate basic hydrograph characteristics; and (ii) the updated modelling platform allows for the explicit assessment of the impact of wetlands (e.g. typology and location) on watershed hydrology. Copyright © 2015 John Wiley & Sons, Ltd.

Simulation and verification of hydraulic properties and organic matter degradation in sand filters for greywater treatment.

To evaluate the treatment performance of vertical flow sand filters, the HYDRUS wetland module was used to simulate treatment in an experimental set-up. The laboratory filters were intermittently dosed with artificial greywater at a hydraulic loading rate of 0.032 m³ m⁻² day⁻¹ and an organic loading rate of 0.014 kg BOD5 m⁻² day⁻¹. The hydraulic properties of the filter were characterised, as were inflow and outflow concentrations of chemical oxygen demand (COD), biochemical oxygen demand (BOD), ammonia, nitrate and total nitrogen. The inverse simulation function of the HYDRUS software was used to calibrate the water flow model. The observed effect of water flowing faster along the column wall was included in the inverse simulations. The biokinetic model was calibrated by fitting heterotrophic biomass growth to measurements of potential respiration rate. Emphasis was put on simulating outflow concentrations of organic pollutants. The simulations were conducted using three models of varying degree of calibration effort and output accuracy. The effluent concentration was 245 mg COD L⁻¹ for the laboratory filters, 134 mg COD mg L⁻¹ for the model excluding wall flow effects and 338 mg COD mg L⁻¹ for the model including wall flow effects.