Subsurface Flow Constructed Wetland System Research Articles

Phosphorus removal in laboratory-scale unvegetated vertical subsurface flow constructed wetland systems using alum sludge as main substrate

This research has two eventual goals: (1) To optimize performance of subsurface constructed wetlands for removal of phosphorus (P) (2) To demonstrate that dewatered alum sludge (a by-product), can be reused as a constructed wetland substrate. To achieve these, alum sludge from a water treatment plant was characterized and used as main substrate in four experimental vertical sub-surface flow constructed wetland systems treating dairy farm wastewater. Results show that the alum sludge has suitable hydraulic characteristics (uniformity coefficient = 3.6) for use as a substrate, and in the batch studies, up to 48.6 mg-P was removed by 1 g of the alum sludge at a P concentration of 360 mg-P/l and a dosage of 5 g/l. Results from the experimental systems highlight the significant P removal ability of the alum sludge. However, the inclusion of pea gravel at the infiltrative surface of some of the systems had a negative effect on the P removal performance. Sequential P-fractionation results show that there was no significant increase in the easily extractable P, but for total P, there was significant increase, although this was found to decrease with depth. This study shows that the novel use of dewatered alum sludge can bring about high P removal in vertical subsurface flow constructed wetland systems.

Comparison of horizontal and vertical constructed wetland systems for landfill leachate treatment

The main purpose of this study was to treat organic pollution, ammonia and heavy metals present in landfill leachate by the use of constructed wetland systems and to quantify the effect of feeding mode. The effect of different bedding material (gravel and zeolite surface) was also investigated. A pilot-scale study was conducted on subsurface flow constructed wetland systems operated in vertical and horizontal mode. Two vertical systems differed from each other with their bedding material. The systems were planted with cattail ( Typha latifolia) and operated identically at a flow rate of 10 l/day and hydraulic retention times of 11.8 and 12.5 day in vertical 1, vertical 2 and horizontal systems, respectively. Concentration based average removal efficiencies for VF1, VF2 and HF were NH 4–N, 62.3%, 48.9% and 38.3%; COD, 27.3%, 30.6% and 35.7%; PO 4–P, 52.6%, 51.9% and 46.7%; Fe(III), 21%, 40% and 17%, respectively. Better NH 4–N removal performance was observed in the vertical system with zeolite layer than that of the vertical 2 and horizontal system. In contrast, horizontal system was more effective in COD removal.

AN EXPERIMENTAL THREE-STAGE HYBRID CONSTRUCTED WETLAND SYSTEM FOR REMOVAL OF ORGANICS AND NUTRIENTS FROM DOMESTIC WASTEWATER

An experimental three-stage subsurface flow constructed wetland system was investigated for the removal of nutrient and organic matter from primary-treated domestic wastewater. Built in the Turkey Scientific and Technical Research Council-Marmara Research Center, Gebze, Turkey (Istanbul), the system consisted of a vertical flow-gravel filtration in the first stage, a horizontal-subsurface flow bed vegetated with Iris in the second stage, and a vertical-subsurface flow bed vegetated Phragmites in the third stage. Wastewater was recycled from the vertical bed to the horizontal one at a ratio of 1:2 and 1:1. The study aimed at evaluating the effects of the loading rates and seasonal changes, effects of recycling and rapid drainage, plant uptake, and the effectiveness of the system in the removal of organic matter and nutrients. The average removal efficiencies were high in summer periods. The results show that the average removal efficiencies increased depending on decrease in hydraulic loading rate and increase in recycle ratio. The relationships between the loading rate and the removal rate were linear.

Study of atrazine degradation in subsurface flow constructed wetland under different salinity

To evaluate the treatment capability of subsurface flow constructed wetland (SFCW) and the effect of salinity on the degradation of atrazine, the degradation of atrazine in SFCW was studied. Under the static condition, the degradation of atrazine in SFCW followed first-order kinetics: c = 0.09679 exp(−0.0396 t) ( c, residue concentration, mg l −1; t, retention time, d), with a half-life of approximately 17.5 days. The atrazine degradation kinetic functions were established for salinities of 1.5, 3.0, 5.0, 10.0 and 15.0 g l −1, respectively, which appeared to approach first-order kinetics. The effect of salinity on the atrazine treatment efficiency showed an exponential inhibition: ln k = 3.204 + 0.04991 C ( k, degradation constant; C, NaCl concentration, mg l −1). The attenuation of atrazine in SFCW cannot be a result of hydrolysis or sorption process. It was considered that some bacteria in the wetland system degraded atrazine into deethylatrazine (DEA) and deisopropylatrazine (DIA) and sequentially into CO 2 and H 2O. Salinity impacted on the growth of bacteria resulting in a switch of the microbial community. With the increase of salinity, Shannon–Wiener Diversity Index in the SFCW system declined. The relationship between atrazine degradation constant ( k) and Shannon Index was established as shown in linear phase, y = −0.07286 + 0.0363 x. The positive correlation between them indicated that microbial community played an important role in the atrazine degradation process.

Assessment of performance characteristics of some constructed wetlands in Nepal

Surface water pollution is one of the serious environmental problems in urban centers ofNepal due to the discharge of untreated wastewater. Almost all conventional type wastewaterplants are not functioning well and practically not sustainable due to lack of regularmaintenance and high operation cost. To overcome this, Constructed Wetland (CW)technology for wastewater treatment arises as alternative technology in Nepal. Now, there are12 sub-surface flow CW systems in operation all over the country.This study aims to checking the efficiency and performance evaluation of three CW located atvarious locations in Nepal. The performance evaluation was conducted through collection ofwastewater samples from inlet and outlet sources of treatment plants and measuringsignificant water qualities determining physical and chemical parameters such as BiochemicalOxygen Demand (BOD), Chemical Oxygen Demand (COD), Hydrogen ion concentration(pH), Conductivity, Phosphorus, Total Suspended Solids (TSS), Total Dissolved Solids (TDS)and Total Solids (TS).The CW systems were found efficiently removing and reducing different pollutionparameters. Performance of CW system were excellent for Kathmandu University (KU) andMalpi International School (Malpi) where the efficiency levels ranged between 82% to 96%for BOD, 68% to 85% for COD, 91 % to 99% for TSS, 60% to 83% for TDS and for TS 84%to 92%. The performance efficiency was relatively lower for Thimi Community ScaleWastewater Treatment Plant (Thimi) where the efficiency levels was 56% for BOD, 64% forCOD, 17% for Conductivity, 88% for TSS, 59% for TDS and 80% for TS. The results showthat, the CW systems are useful and efficient for waste water treatment in growing urbanareas. They are cost effective and easy to maintain. It is recommended that CW system couldbe replicate in other growing urban areas.

Research on Nitrogen Removal and Microorganism in a Subsurface Flow Constructed Wetland System in Sihong County

Abstract Experiments in monitoring the removal of organic material and nitrogen and determining the amounts of microorganism at different sites in the subsurface flow constructed wetland in Sihong county were performed. The results show that the removal of CODcr agrees with the kinetic equation of a first order reaction. The removal of pollutants varies with different seasons. The removal rates of CODCr, NH3-N, TN in the spring are 15%–23% higher than those in the autumn. The amount of ammonifier is larger than that of denitrifying bacteria and the amount of denitrifying bacteria is larger than that of nitrosomonas. The amount of bacteria around the plant roots is larger than that on the surface of the packing medium. No apparent change is observed for the amount of denitrifying bacteria and nitrosomonas between spring and autumn.

Cadmium Removal from Wastewater in Constructed Wetlands

This study was conducted to investigate the cadmium removal in the free water surface and subsurface flow constructed wetland systems under different environmental conditions. Removal efficiencies of the chemical oxygen demand were found to be in the range of 79–92% for the two wetland systems. The removal efficiencies for total kjeldahl nitrogen, total phosphorus, total suspended solid, and volatile suspended solid were in the range of 64–91, 62–90, 68–91, and 50–84%, respectively. Influents were obtained by mixing a synthetic wastewater with Cd Cl2 . H2 O at four concentrations 1, 5, 10, and 20 mg∕L , respectively, for the four experimental runs. The mean effluent cadmium concentrations varied between 0.02 and 0.16 mg∕L . Cadmium removal efficiencies were 98.6–99.4% in free water surface, and 99.3–99.9% in the subsurface flow wetland systems. The subsurface flow wetland showed slightly better performance than the free water surface wetland in terms of various pollutants’ removal. However, the two system...

Nitrogen Transformations Modeling in Subsurface‐Flow Constructed Wetlands

Subsurface-flow constructed wetlands (CWs) wastewater treatment typically results in satisfactory organics removal. However, the removal of nutrients, particularly nitrogen, is often unreliable, and typically less than desired, and nitrogen transformations in wetlands systems are not well-understood. The principal objective of this study was to establish a basis for quantification of nitrogen transformations through subsurface flow CW systems. Actual performance data from a full-scale facility located near Lincoln, Nebraska, were used to calibrate a proposed nitrogen transformations model, which, in turn, was used to replicate and predict the wetlands performance. To realize this objective, a compartmental analysis technique, which uses a set of differential equations and nonlinear optimization numerical methods, was used for solving nitrogen transformation rates and for predicting wetland performance. The model satisfactorily reproduced the mean effluent concentrations for organic nitrogen, ammonium-nitrogen, and nitrate-nitrogen, but with lesser accuracy with respect to peak high and low effluent concentrations. Nitrogen mass balance in the wetland was used to identify likely nitrogen transformation pathways. Generally, it was found that approximately one-third of the influent nitrogen mass was removed through nitrification and denitrification, one-third was removed through vegetative assimilation, and the remainder was discharged in the wetland effluent.

Estimating Inorganic and Organic Nitrogen Transformation Rates in a Model of a Constructed Wetland Purification System for Dilute Farm Effluents

This paper describes some of the concepts which are being built into a model of a subsurface flow constructed wetland system based on reedbeds. A nitrogen (N) cycling submodel has been adapted from a soil N model, with pools representing organic material with high biological oxygen demand (BOD), ammonium and nitrate. Microbiologically controlled transformations between pools are represented by first-order exponential kinetics, with N finally lost to the atmosphere either by ammonia volatilisation, or by denitrification to gaseous N 2 or nitrous oxide. Hydrology has been represented for both horizontal and vertical reedbeds, with contrasting partially anaerobic or fully aerobic conditions at each stage. The model has been set up to represent an experimental system with one horizontal and three vertical reedbeds, being tested for purification of dirty water from a dairy farm. Rate constant values have been selected so that simulated results are a reasonable approximation to measurements from the experimental system. The model has been developed to assist in optimising design parameters for new systems, including the numbers and sequence of reedbed types, dimensions and flowrates, for various incoming contaminant concentrations and target water purity standards.

Nitrogen transformation in horizontal subsurface flow constructed wetlands I: Model development

In this paper a mathematical model for prediction of nitrogen transformation in horizontal subsurface flow constructed wetlands was developed. Two horizontal subsurface flow constructed wetlands were designed to receive organic loading rate below 50 kg/ha/d and hydraulic loading rate of 480 m 3/ha/d from a primary facultative pond. Two rectangular shaped units each 11.0 m long, 3.7 m wide and 1.0 m deep and bottom slope of 1% were constructed and filled with 6–25 mm diameter gravel pack to a depth of 0.75 m. Each unit was planted with Phragmites mauritianus with an initial plant density of 29,000 plants/ha. The plants were allowed to grow for about four months before sampling for water quality parameters commenced. Samples were collected daily for about three months. Dissolved oxygen, pH and temperature were measured in situ and ammonia, total Kjeldahl nitrogen, nitrates, nitrite and Chemical Oxygen Demand were measured in the laboratory in accordance with Standard Methods. The mathematical model took into account activities of biomass suspended in the water body and biofilm on aggregates and plant roots. The state variables modelled include organic, ammonia, and nitrate–nitrogen, which were sectored in water, plant and aggregates. The major nitrogen transformation processes considered in this study were mineralization, nitrification, denitrification, plant uptake, plant decaying, and sedimentation. The forcing functions, which were considered in the model, are temperature, pH and dissolved oxygen. Stella II software was used to simulate the nitrogen processes influencing the removal of nitrogen in the constructed wetland. One of the two-wetland units was used for model calibration and the second unit for model validation. The model results indicated that 0.872 gN/m 2 d was settled at the bottom of the wetland and on gravel bed and roots of the plants. However, 0.752 gN/m 2 d (86.2%) of the settled nitrogen was regenerated back to the water body, which means that only 13.8% of the settled nitrogen was permanently removed. Denitrification and nitrification were responsible for transformation of 0.436 gN/m 2 d and 0.425 gN/m 2 d, respectively. Uptake of nitrogen by plants was 0.297 gN/m 2 d out of which 0.140 gN/m 2 d was returned to the water body as plants decay. It was found that the major pathways leading to permanent removal of nitrogen in a horizontal subsurface flow constructed wetland system in descending order are denitrification (29.9%), plant uptake (10.2%) and net sedimentation (8.2%). A total nitrogen removal of 48.9% was achieved in this study.

Simazine and Metolachlor Removal by Subsurface Flow Constructed Wetlands

A subsurface flow constructed wetlands system was used in this pilot-scale field study to treat runoff from a container nursery. The goal of the study was to determine the statistical significance of certain factors on the removal of simazine (2-chloro-4,6-bis[ethylamino]-s-triazine) and metolachlor (2-chloro-N-[2-ethyl-6-methylphenyl]-N-[2-methoxy-1-methylethyl] acetamide). The factors analyzed included the hydraulic flowrate, media depth, cell aspect ratio, and the presence or absence of Scirpus validus. Analysis of variance using the general linear model procedure determined that vegetation and flowrate significantly (alpha = 0.05) affected herbicide mass removal. The results showed that vegetated subsurface flow cells generally remove more simazine and metolachlor than nonvegetated cells. Decreasing the hydraulic loading rate and, therefore, the areal simazine mass loading rate to vegetated cells from 1659 to 415 mg/m2 caused an increase in the percentage of total mass removed from approximately 60 to 96%. Similarly, metolachlor mass removal percentage increased from approximately 62 to 96% when the areal mass loading to vegetated cells decreased from 1037 to 260 mg/m2. Media depth and aspect ratio did not significantly (alpha = 0.05) affect effluent simazine and metolachlor concentrations. Peak effluent metolachlor concentrations were less than 300 microg/L from vegetated cells receiving hydraulic loading rates of 204 and 102 m3/d x ha. Peak simazine concentrations in the effluent of cells receiving similar hydraulic loadings were approximately 1400 and 750 microg/L (204 and 102 m3/d x ha), respectively.

Design methodology and area sensitivity analysis of horizontal subsurface flow constructed wetlands

An approach is presented to satisfy the demand for simple criteria, guidelines and models for the preliminary sizing of horizontal subsurface flow (SF) constructed wetland systems. This approach eliminates time-consuming calculations and iterations by providing graphical solutions for wetland system sizing. Therefore, it can be used for the preliminary assessment of new or performance evaluation of existing subsurface flow constructed wetland systems. The validity of this methodology is checked with data from existing systems and is found to be quite satisfactory. This methodology is combined with simple equations predicting the maximum wetland capacity in summer, so as to assist designers in sizing installations in tourist areas with increased summer populations. Furthermore, based on this methodology, a sensitivity analysis is performed of the area requirements for wastewaters of various strengths and various design conditions and performance criteria. The results provide a useful overview to engineers and further simplify the design methodology of new subsurface flow constructed wetland systems.

Removal of nutrients from combined sewer overflows and lake water in a vertical-flow constructed wetland system

Lake Utterslev is situated in a densely built-up area of Copenhagen, and is heavily eutrophicated from combined sewer overflows. At the same time the lake suffers from lack of water. Therefore, a 5,000 m2 vertical flow wetland system was constructed in 1998 to reduce the phosphorus discharge from combined sewer overflows without reducing the water supply to the lake. During dry periods the constructed wetland is used to remove phosphorus from the lake water. The system is designed as a 90 m diameter circular bed with a bed depth of c. 2 m. The system is isolated from the surroundings by a polyethylene membrane. The bed medium consists of a mixture of gravel and crushed marble, which has a high binding capacity for phosphorus. The bed is located within the natural littoral zone of the lake and is planted with common reed (Phragmites australis). The constructed wetland is intermittently loaded with combined sewer overflow water or lake water and, after percolation through the bed medium, the water is collected in a network of drainage pipes at the bottom of the bed and pumped to the lake. The fully automated loading cycle results in alternating wet and dry periods. During the initial two years of operation, the phosphorus removal for combined sewer overflows has been consistently high (94-99% of inflow concentrations). When loaded with lake water, the phosphorus removal has been high during summer (71-97%) and lower during winter (53-75%) partly because of lower inlet concentrations. Effluent phosphorus concentrations are consistently low (0.03-0.04 mg/L). Ammonium nitrogen is nitrified in the constructed wetland, and total suspended solids and COD are generally reduced to concentrations below 5 mg/L and 25 mg/L, respectively. The study documents that a subsurface flow constructed wetland system can be designed and operated to effectively remove phosphorus and other pollutants from combined sewer overflows and eutrophicated lake water.

Media selection for sustainable phosphorus removal in subsurface flow constructed wetlands

Sorption of phosphorus (P) to the bed sand medium is a major removal mechanism for P in subsurface flow constructed wetlands. Selecting a sand medium with a high P-sorption capacity is therefore important to obtain a sustained P-removal. The P-removal capacities of 13 Danish sands were evaluated and related to their physico-chemical characteristics. The P-removal properties of sands of different geographical origin varied considerably and the suitability of the sands for use as media in constructed reed beds thus differs. The P-sorption capacity of some sands would be used up after only a few months in full-scale systems, whereas that of others would subsist for a much longer time. The most important characteristic of the sands determining their P-sorption capacity was their Ca-content. Also the P-binding capacities of various artificial media were tested (light-expanded-clay-aggregates (LECA), crushed marble, diatomaceous earth, vermiculite and calcite). Particularly calcite and crushed marble were found to have high P-binding capacities. It is suggested that mixing one of these materials into the sand or gravel medium can significantly enhance the P-sorption capacity of the bed medium in a subsurface-flow constructed wetland system. It is also possible to construct a separate unit containing one of these artificial media. The media may then be replaced when the P-binding capacity is used up.

Subsurface-flow constructed wetlands treatment in the plains: five years of experience

This paper documents the performance of a subsurface-flow constructed wetlands system during its initial five years of operation under variable loading and operating conditions associated with a northern midwestern US climate. The results indicate that effective and sufficient CW seasonal removals of TSS, VSS, CBOD5, COD, and fecal coliform were achieved. Wastewater temperatures seemed to affect CBOD5 and COD removal rates. Nitrogen and phosphorus reductions were not as effective and varied seasonally, as well as with wastewater temperature. The addition of a sand filter, to aid in further nitrification and disinfection following CW treatment, markedly improved the performance of the wetlands system. After a few years of operation, the remarkable performance of the CW system was dampened by apparent clogging and subsequent eruption of wastewater at the head-end of the treatment cells. While clogging was partially caused by biomass build-up in the wetlands substrate, visual observations suggest that excessive vegetation coupled with relaxed maintenance may also be responsible for clogging.

Performance modeling of subsurface-flow constructed wetlands systems

A subsurface flow constructed wetlands (CW) system, located at a neighborhood consisting of a small housing development and golf courses outside of Lincoln, NE, was studied for its effectiveness as a small community wastewater system. Extensive monitoring was conducted biweekly between June 1996 and December 2000. Prediction models for soluble CBOD5 NH3-N, and TP removal in CW were employed for comparison with the field data. It was found that the disappearance of BOD5 and NH3-N could be approximated using first-order kinetics, but the kinetics of TP removal were unclear. The reduction rate constants regressed from the field data were found to be lower than literature reported values.

Pathogen and Pathogen Indicator Removal Characteristics in Treatment Wetlands Systems

This paper reports the performance of two horizontal subsurface flow constructed wetland systems in the removal of pathogen and pathogen indicators under variable loading and operating conditions. The two constructed wetland systems evaluated under this project are located near Lincoln, Nebr. Firethorn is located in a small housing community and Rogers Farm at a rural single-family dwelling. Samples at Firethorn were collected at the wetland cell influent, effluent, and sand filter effluent (system effluent) from June 1996 through April 2000 and at the influent and effluent points from Rogers Farm from June 1999 through April 2000. The main monitoring parameters were fecal coliform at both sites and Salmonella spp. (from August 1999 through January 2000) at Firethorn. Firethorn demonstrated effective removal of fecal coliforms with 96.3% removal in the wetland cell and 98% by the wetland system, which included a sand filter. Rogers Farm had an average fecal coliform removal of 99.3%. Although Salmonella spp. was not detected at Firethorn, reduction of other bacteria was observed through the wetland system. In general, fecal coliform removal was excellent through the wetland system and no distinct difference in removal efficiency was observed with changes in season and temperature at either facility.

Phosphorus retention capacity of root bed media of sub-surface flow constructed wetlands

Phosphorus retention by sub-surface flow constructed wetlands is dependent upon the effluent quality, loading rate and type of root bed media. Three types of root bed media (Lockport dolomite, Queenston shale and Fonthill sand) at various stages of their use were sampled from a sub-surface flow wetland located in Sewage Waste Amendment Marsh Process Project (SWAMP), Niagara-on-the-Lake, Ontario, Canada, and their P sorption characteristics were investigated. Lower equilibrium P concentration (EPC o) and higher P sorption maxima (S max) and retention capacity (P r) of untreated Fonthill sand compared to untreated Lockport dolomite and Queenston shale indicated that Fonthill sand could be better root bed media for sub-surface flow constructed wetland systems to remove P from wastewater. In general, EPC o of the root bed media increased with the duration of their use. Untreated root bed media indicated that the greater the amount of poorly crystalline forms of aluminum (Al), and magnesium (Mg) contents, the greater their EPC 0. The P sorption maxima of root bed media increased with use by accumulation of amorphous and poorly crystalline forms of Al and iron (Fe). However, in alkaline root bed media, preferential P sorption occurred more on amorphous and poorly crystalline forms of Al than Fe. This study also indicated that increases in P sorption capacities of root bed media do not necessarily assure lower effluent P concentrations, since EPC o increased by several fold with use.

Subsurface Flow Constructed Wetlands Performance Evaluation Using Area-Based First-Order Kinetics

Subsurface Flow Constructed Wetlands Performance Evaluation Using Area-Based First-Order KineticsThis paper discusses the effectiveness of commonly used area-based first-order reduction models in predicting the performance of constructed wetland (CW) system. A subsurface flow CW systems in a small community near Lincoln, Nebraska, has been studied as a natural treatment system to evaluate several design approaches used in sizing and predicting the performance of CW systems. Extensive...Author(s)Wenxin LiuM.F. DahabR.Y. SurampalliSourceProceedings of the Water Environment FederationSubjectSession 14 - Natural Systems, Water Reuse and Small Communities: Wetland Systems in Small CommunitiesDocument typeConference PaperPublisherWater Environment FederationPrint publication date Jan, 2000ISSN1938-6478SICI1938-6478(20000101)2000:13L.272;1-DOI10.2175/193864700784608027Volume / Issue2000 / 13Content sourceWEFTECFirst / last page(s)272 - 282Copyright2000Word count181

Subsurface flow constructed wetlands treating municipal wastewater for nitrogen transformation and removal

Research was conducted on pilot‐scale, subsurface flow constructed wetland systems treating municipal wastewater. Increasing the hydraulic retention time (HRT) from 1.7 days to 4.3 days reduced the mean effluent NH4 concentration by 2.7 mg N/L. Hydroperiod manipulation caused a 2.9‐mg N/L decrease in effluent NH4. The influent NH4 concentration for the 9‐month study period averaged 21 mg N/L. The lowest average effluent NH4 concentration observed in single‐stage treatment cells was 13 mg N/L. The mean effluent NH4 concentration from the second‐stage cells of two wetland cells in series was 10 mg N/L. During the period from May through September, the second‐stage cells produced an effluent NH4 concentration of 6 mg N/L. A variable‐order kinetic model for sizing the second‐stage cells of a series system yielded a coefficient of determination, r2 of 0.94 for the summer NH4 removal data. The predicted HRT to reduce NH4 from 20 mg N/L to 5 mg N/L was 3.9 days, if the cells were subjected to cyclic draw and fill.