DeBusk, Thomas A. et al “Wetlands for Water Treatment” Applied Wetlands Science and Technology Editor Donald M. Kent Boca Raton: CRC Press LLC,2001 ©2001 CRC Press LLC CHAPTER 9 Wetlands for Water Treatment Thomas A. DeBusk and William F. DeBusk CONTENTS General Features of Wetlands that Contribute to Contaminant Removal Types of Treatment Wetlands Free Water Surface Wetlands Subsurface Flow Wetlands Hybrid Treatment Wetlands Treatment Wetland Components Treatment Wetland Vegetation Hydroperiod and Hydraulics Treatment Wetland Soils Treatment Wetland Contaminant Removal Processes Physical Removal Processes Biological Removal Processes Chemical Removal Processes Planning and Design Treatment Wetlands as a Unit Process Regulatory Issues Preliminary Feasibility and Alternatives Analyses Design Considerations Construction and Management Construction Management Performance Suspended Solids and Organic Carbon Removal Removal of Organic Carbon and Suspended Solids from Waste Stabilization Pond Effluents ©2001 CRC Press LLC Nitrogen Removal Nitrogen Removal from Food Processing Wastewaters Phosphorus Removal Removal of Trace Metals, Toxic Organic Compounds, and Complex Mixtures of Contaminants Wetland Treatment of Urban Runoff Wetland Treatment of Landfill Leachates Pathogen Removal Conclusions References Wetlands are widely regarded as biological filters, providing protection for water resources such as streams, lakes, estuaries, and groundwater. Although naturally occurring wetlands have always served as ecological buffers, research and develop- ment of wetland treatment technology is a relatively recent phenomenon. Studies of the feasibility of using wetlands for wastewater treatment were initiated during the early 1950s in Germany. In the United States, wastewater to wetlands research began in the late 1960s and increased dramatically in scope during the 1970s. As a result, the use of wetlands for water and wastewater treatment has gained considerable popularity worldwide. Currently, an estimated 1000 wetland treatment systems, both natural and constructed, are in use in North America (Cole, 1998). The goal of water and wastewater treatment is the removal of aqueous contam- inants in order to decrease the possibility of detrimental impacts on humans and the rest of the ecosystem. The term contaminant is used in this context to refer to any constituent in the water or wastewater that may adversely affect human and envi- ronmental health. Many contaminants, including a wide variety of organic com- pounds and metals, are toxic to humans and other organisms. Other types of con- taminants may not be hazardous in the conventional sense but nevertheless pose an indirect threat to our well being. For example, loading of nutrients (e.g., nitrogen and phosphorus) to waterways can result in excessive growth of algae and unwanted vegetation. This growth diminishes the recreational, economic, and aesthetic values of lakes, bays, and streams. Constructed wetlands have been successfully used as treatment systems for domestic wastewater effluent, from single-residence wetlands to large municipal wastewater treatment facilities. Similarly, wetlands may be used effectively for treatment of animal and aquaculture wastes. The use of wetland retention basins for treatment of stormwater runoff has become relatively commonplace. The composi- tion of stormwater varies greatly, depending on the surrounding land use. For exam- ple, urban runoff may contain soil particles, dissolved nutrients, heavy metals, oil, and grease. Residential and agricultural runoff may also contain organic matter and pesticides. A variety of industrial wastes, including pulp and paper, food processing, slaughtering and rendering, chemical manufacturing, petroleum refining, and landfill leachates are amenable to wetland treatment. ©2001 CRC Press LLC While wetlands are remarkable in their ability to treat diverse types of contam- inated waters, there are limitations that should be carefully considered prior to the implementation of any wetland treatment system. Pretreatment, for example, primary sedimentation or anaerobic or aerobic stabilization, is often required prior to feeding domestic and industrial effluents into a wetland. Selected water contaminants removed within a treatment wetland ultimately may become available for assimila- tion and potentially be toxic to wetland biota. A sound understanding of wetland contaminant removal processes (Reddy and D’Angelo, 1997), the long-term fate of these contaminants, and contaminant removal effectiveness of various wetland types is critical in the proper design and operation of treatment wetlands. GENERAL FEATURES OF WETLANDS THAT CONTRIBUTE TO CONTAMINANT REMOVAL The unique combination of structural and functional attributes sets wetlands apart from terrestrial and aquatic ecosystems in their ability to remove or sequester nutrients and toxic environmental contaminants. For example, shallow water, low current velocity, and the physical filtering action of plant stems and leaves provide favorable conditions for settling of particulate matter. Wetlands also provide sub- strates for a multitude of chemical and microbiological processes, promoting nutrient removal and storage within the complex maze of microsites in the soil and vegetation cover. The total surface area available for microbial activity in the soil and the overlying dead plant material (litter or detritus) is extremely high in wetlands. Physical, chemical, and microbiological processes are further enhanced in wetlands by retention of water for extended periods within this biologically active zone. Another important characteristic of wetlands is the presence of anaerobic (oxygen- depleted) soils during periods of flooding which gives rise to an aerobic–anaerobic interface, or boundary, near the soil surface. This juxtaposition of aerobic and anaerobic conditions provides an environment for unique chemical and microbio- logical reactions that greatly enhance the removal of nutrients from inflowing water. Wastewater-borne labile organic carbon compounds, expressed as biochemical oxygen demand (BOD), are readily removed in anaerobic and aerobic microenvi- ronments of treatment wetlands. Reduced nitrogen (N) compounds (e.g., ammonium) are nitrified in aerobic regions, from which the products can migrate (either by bulk transfer or diffusion) to anaerobic regions. Denitrification of the produced NO x species occurs rapidly due to the preponderance of anaerobic conditions and ready availability of labile carbon compounds from decaying vegetation and organic soils. In this sequential N removal process, nitrification typically is the rate-limiting step due to low oxygen availability in many parts of the system. Plant uptake and adsorption to soil surfaces contribute to short-term phosphorus (P) removal in wetlands. However, the only prominent, long-term P sink is thought to occur through soil accumulation. Large wetland areas are therefore required to achieve substantial P removal. Treatment wetlands differ in two fundamental ways from more conventional wastewater treatment unit processes. First, wetlands sacrifice consistently high ©2001 CRC Press LLC microorganism densities and strict process control for reduced construction costs and operator attention. Second, solids processing occurs internally in wetlands, so no biological sludge management is required, at least on the short-term. For the above reasons, treatment wetlands have moderate to high land requirements. In summary, a number of physical, chemical, and biological processes operate concurrently in constructed and natural wetlands to provide contaminant removal (Figure 1). Removal of contaminants may be accomplished through storage in the wetland soil and vegetation or through losses to the atmosphere. Knowledge of the basic contaminant removal concepts is extremely helpful for assessing the potential applications, benefits, and limitations of wetland treatment systems. These processes are described in more detail in a later section. TYPES OF TREATMENT WETLANDS Treatment wetlands are generally classified as either free water surface (FWS) or subsurface flow (SSF) systems (Figure 2). Subsurface flow wetlands are the common system design implemented in Europe for domestic wastewater treatment which has greater than 500 treatment wetlands. In North America, with around 600 treatment wetlands, the FWS type is more common (Cole, 1998). In the United States, FWS wetlands for domestic wastewater treatment commonly occur in communities with 1000 or fewer people, although some large FWS wetlands exist in cities with popu- lations greater than 1 million. As of late 1998, South Dakota was the state with the greatest number of operational (nonpilot) FWS wetlands (42), followed by Florida (24) and California (11) (U.S. Environmental Protection Agency, 1999). The widespread use of treatment wetlands in South Dakota, a state with harsh winter conditions, provides a good indication of the versatility of treatment wetlands, as well as the circumstances under which they are an appropriate and competitive technology. Because of low capital (owing to inexpensive land) and operating costs and the ability to provide winter water storage, waste stabilization ponds (WSPs) have been widely implemented for domestic wastewater treatment during the past four decades. As of 1991, 246 communities in South Dakota were using WSPs. A drawback of WSPs is that they do not consistently provide low effluent suspended solids, ammonium, and total phosphorus concentrations. A strong interest in addi- tional protection of the quality of water resources, as well as in creating new wildlife habitat, has led to the upgrading of many WSPs with treatment wetlands during the past decade (Dornbush, 1993). The South Dakota state regulatory agency has encour- aged the use of constructed wetlands by providing design guidelines and economic assistance with treatment wetland construction. Free Water Surface Wetlands FWS design typically incorporates a shallow layer of surface water, flowing over mineral (sandy) or organic (peat) soils. Vegetation often consists of marsh plants, such as Typha (cattails) and Scirpus (bulrush), but may also include floating and submerged aquatic vegetation and wetland shrubs and trees (Figure 3). Natural ©2001 CRC Press LLC Figure 1 Contaminant removal mechanisms and transformations in free water surface (FWS) wetlands. Volatilization NH3 Volatile organics Plant storage Sedimentation Plant uptake Inflow CO2, CH4 Decomposition Organic C Burial Soil storage (peat) N2 Denitrification NO3- Adsorption Precipitation - NH4+, metals, P, organics (to clays, Fe/Al hydroxides, organic matter) - P (with Fe, Al, Ca) - Metals (with sulfides) SURFACE WATER DETRITUS (LITTER) SOIL ©2001 CRC Press LLC wetlands, both forested and herbaceous, have also been effectively used as FWS treatment wetlands. For some treatment applications, FWS wetlands are designed and managed to encourage dominance by either floating or submerged macrophytes. Water depth is one parameter that can be controlled to discourage emergent macrophytes, thereby allowing development of either a floating aquatic macrophyte (FAM) or submerged aquatic vegetation (SAV) system. Free water surface wetlands vary dramatically in size, from less than 1 ha to greater than 1000 ha. Large FWS wetlands are even being used as a nutrient control technology to treat runoff from entire regional watersheds. For example, over 16,000 ha of FWS wetlands are being constructed in South Florida to remove P from agricultural drainage water before it enters the Everglades (Moustafa et al., 1999). Free water surface wetlands offer ecological and engineering benefits beyond water treatment. Free water surface wetlands used for treating agricultural and urban Figure 2 Schematic of free water surface (FWS) and subsurface flow (SSF) wetlands. FREE WATER SURFACE WETLAND SUBSURFACE-FLOW WETLAND Inflow Inflow Outflow Outflow SURFACE WATER DETRITUS (LITTER) SOIL (MINERAL OR PEAT) WATER LEVEL DETRITUS (LITTER) GRAVEL OR SOIL ©2001 CRC Press LLC runoff also reduce hydraulic runoff peaks from storm events. Effectiveness depends on the wetland size (volume), location in the watershed, and configuration of inlet and outlet structures (Schueler, 1996). Many FWS treatment wetlands provide both a recreational amenity and wildlife habitat. Iron Bridge wetland in Orlando, FL and the Arcata marsh in Humbolt, CA have each provided water treatment and other ecological and aesthetic benefits for more than a decade (Jackson, 1989; Gearhart and Higley, 1993). Subsurface Flow Wetlands Subsurface flow wetlands differ from FWS wetlands in that they incorporate a rock or gravel matrix that the wastewater is passed through in a horizontal or vertical fashion (see Figure 2). Unless the matrix clogs, the top layer of the bed in horizontal flow systems will remain dry. The SSF configuration offers several advantages, including a decreased likelihood of odor production and no insect proliferation within the wetland as long as surface ponding is avoided. Unlike FWS wetlands, SSF systems provide no aesthetic or recreational benefits and few, if any, benefits to wildlife. Subsurface flow wetlands continue to provide effective treatment of most waste- water constituents through the winter in temperate climates. The subsurface micro- bial treatment processes still function, albeit at a reduced rate, even when the surface vegetation has senesced or died, and the matrix surface is covered with snow and ice. Subsurface flow wetlands also can be operated in a vertical flow fashion which Figure 3 Free water surface (FWS) wetlands incorporate a shallow layer of surface water and vegetation such as these emergent macrophytes. ©2001 CRC Press LLC can reduce matrix clogging problems and enhance certain contaminant removal processes such as nitrification. Because of the high cost of the gravel or rock matrix, SSF wetlands never attain the large spatial footprint of the large FWS wetlands. Concerns over matrix clogging and the potential high cost of renovation also limit the deployment of extremely large SSF wetlands. However, SSF are finding increased use for small applications, such as for small communities or single family homes. The limitations of septic systems for nutrient control have become more apparent in the past two decades (Hagedorn et al., 1981), and SSF wetlands are one technology that is being deployed to improve nutrient removal performance (Mitchell et al., 1990). Subsurface flow systems are the only wetland configuration suitable for this purpose, because they create no standing water, thereby limiting the likelihood of human exposure to wastewater pathogens (House et al., 1999). Hybrid Treatment Wetlands A number of treatment wetlands have been constructed that combine different wetland types. Some, such as the 134-ha Eastern Service Area wetland in Orlando, FL, consist of constructed FWS wetlands that are followed by natural forested wetlands. This particular configuration was based on regulatory needs, with the natural parcel receiving water only after pretreatment by the constructed wetland (Schwartz et al., 1994). Other hybrid systems are based on specific contaminant removal needs. For example, the key to enhanced nitrogen removal in SSF wetlands is to create an intermediate step in the process train with an oxygenated environment that enhances nitrification in the rock or gravel matrix. Many subsurface wetlands accomplish this by sequencing horizontal flow beds with vertical flow beds or by recirculating partially treated effluent onto an inflow region rock or sand filter to enhance nitrification (Cooper et al., 1997; Reed and Brown, 1995). Other investiga- tors have recommended operating vertical flow beds with various draw and fill cycles to enhance chemical oxygen demand (COD) and nitrogen removal, but performance using this technique has been mixed (Burgoon et al., 1995; Boutin et al., 1997). Regardless of the approach used to stimulate nitrification, the SSF wetland is con- figured so that the nitrate-rich effluent subsequently is introduced into an anoxic section of the bed where denitrification readily occurs. TREATMENT WETLAND COMPONENTS Treatment Wetland Vegetation Macroscopic vegetation is the most prominent feature of treatment wetlands. Free water surface wetlands can develop as simple monocultures of weedy or competitive species such as Typha (cattail) or Phragmites (reed), but more often they contain a diversity of other emergent and floating plants within genera such as Pontederia, Sagittaria, Eleocharis, Utricularia, and Lemna . Treatment wetlands typically are planted just prior to initial flooding to ensure rapid vegetative cover ©2001 CRC Press LLC development and to facilitate initiation of water treatment. Under extreme operating conditions, such as high organic, nutrient, or hydraulic loading rates, the system may remain a monoculture or near monoculture for the life of the treatment system. Similarly, subsurface flow wetlands usually remain dominated by the species planted prior to startup, typically Phragmites (reed), Scirpus (bulrush), or Typha (cattail). This is due both to the difficulty of seeds and other propagules in becoming established on the bed’s surface and the often high organic loading provided to SSF systems. Under less rigorous environmental conditions, the vegetative community that develops over time in FWS wetlands may bear little resemblance to the species originally planted. At the Eastern Service Area Treatment Wetland (Orlando, FL), one constructed wetland is quite shallow and experiences periodic drydown, and the second, while also shallow, is continuously inundated. This system is used for further polishing of domestic wastewater that has received conventional, advanced treatment to levels below 5 mg BOD/l, 5 mg total suspended solids (TSS)/l, 3 mg N/l, and 1 mg P/l (Schwartz et al., 1994). Upon wetland startup, 13 species were planted into the mineral soils at a density of 336 plants per ha. The vegetative communities in the two constructed wetlands were sampled at 1 and 4 years after planting. Only one of the most abundant species occurring at year four in each wetland was a species that was originally planted (Table 1). In a wetland ecosystem self-organization experiment, Mitsch et al. (1998) estab- lished two individual 1-ha wetlands for treating Olentangy River water in Ohio. One wetland was planted with 2400 propagules (rootstock and rhizomes), representing 13 plant species at an overall density of 0.24 plants/m 2 . Species planted included Nelumbo lutea , Nymphaea odorata, and Potamogeton pectinatus in the deepest (0.6 m depth) region, Scirpus validus and Scirpus fluviatilis at moderate (0.3 m) depth, and Spartina pectinata , Sparganium eurycarpum , Acorus calamus , Sagittaria latifolia , Pontedereria cordata , Juncus effusus , Saururus cernuus, and Cephalanthus occidentalis in the shallow, littoral (0 to 0.3 m) region. The second wetland, adjacent to the first, was left unplanted. The wetlands were evaluated each year after start- up for water treatment aspects and flora and fauna characteristics. As of year three, 9 of the 13 original stocked species in the planted wetland were still present, although the total number of macrophyte species had increased to 65 (Mitsch et al., 1998). The unplanted wetland had similar vegetation at year three, and had 54 macrophyte species. However, only 1 of the 13 species originally planted in the adjacent wetland were present. Treatment performance and fauna of the two systems were remarkably similar at year three. Whether or not to plant a FWS treatment wetland upon start-up, as well as what density to plant, is dictated by the urgency to achieve an operational system. If the system is built a year or two in advance of water or wastewater treatment needs, and the design does not call for any specific vegetation components, then existing studies clearly show that natural recruitment can be relied upon for development of a diverse plant community. Depending on the depth and nutrient regime of the wetland, mats of filamentous algae, phytoplankton, or submerged macrophytes likely will dominate in the wetland water column prior to development of a dense emergent macrophyte community. [...]... Health Organization, 198 9) Studies with various treatment wetland configurations have demonstrated a 1-log (90 percent) to 2-log (99 percent) reduction for indicator bacteria, depending on the system HRT Studies with pilot-scale SSF wetlands planted with Phragmites demonstrated 1. 1- to 1 . 9- log removals of Escherichia coli and total coliforms at HRTs in as short as 6 hours (Green et al., 199 7) These removals... macrophyte-dominated treatment wetlands (DeBusk et al., 199 0; Kadlec and Knight, 199 6) reveal that flow patterns depart widely from ideal plug-flow characteristics Temperature-related water column density gradients, the heterogeneous and clumped nature of vegetation, and uneven microtopographical features result in the development of rapid flow paths and internal dispersion and mixing (Kadlec, 199 0) The... N (inorganic plus organic forms) concentrations in North American wetlands was 9. 03 to 4.27 mg/l for FWS wetlands and 18 .92 to 8.41 mg/l for SSF wetlands Mean total N mass removal rates averaged 1.06 and 5.85 kg/ha per day for FWS and SSF wetlands, respectively (Kadlec and Knight, 199 6) Note that the higher mass removal rates for SSF wetlands than FWS systems for N, BOD5 , and TSS is related principally... the water The ©2001 CRC Press LLC Table 4 Summer Performance (mg/l) of a Pilot-Scale Combined Wetland for Treating Potato Processing Wastewaters* Influent COD TSS Organic N NH4-N NO3-N HSF1 HSF2 298 6 607 91 73 1 1056 85 10 1 29 1 601 72 3 116 1 Effluent VFW3 2 09 48 13 26 43 HSF4 161 37 12 29 13 * The initial two horizontal flow wetlands, HSF1 and HSF2, were designed for chemical oxygen demand (COD) and suspended... From a performance-forecasting standpoint, these deviations from plug-flow have been addressed by using different hydraulic reactor models For example, HRT has been modeled using several continuously stirred tank reactors (CSTRs) in series, or a plug-flow reactor followed by multiple CSTRs (Kadlec, 199 7; King et al., 199 7) Recognition of nonideal flow characteristics, as documented by full-scale tracer studies,... Protection Agency, lists the performance of several hundred treatment wetlands Several authors and agencies have provided overviews of the NAWTSD, as well as discussions of the benefits and limitations of these data for design purposes (Kadlec and Knight, 199 6; U.S Environmental Protection Agency, 199 9) Both area-specific and volume-specific models are available for treatment wetland design purposes Both... species (DeBusk et al., 199 6a; Reddy and DeBusk, 198 5) Among emergent macrophytes, cattail (Typha), bulrush (Scirpus), and reed (Phragmites) provide some of the highest N and P removal rates Short-term P uptake rates in excess of 37 g P/m2 per year have been reported for these productive floating and emergent species (Reddy and DeBusk, 198 5; Tanner, 199 6) Despite obvious between-species differences in... differences ©2001 CRC Press LLC in habit influence water column and sediment-water interface environmental conditions, such as dissolved oxygen concentrations and solar radiation inputs, that are the critical master factors in controlling element cycling and contaminant removal in wetlands (Reddy and D’Angelo, 199 7; Reddy et al., 199 9) Finally, where prominent plant morphological differences exist (or differences... For phosphorus, the k value is low because burial of P in the sediment is thought to be the only long-term P removal process (Richardson and Craft, 199 3) Phosphorus is actually quite dynamic in wetlands, owing to settling of influent particulate P, assimilation and short-term or long-term storage of ortho- or soluble reactive P by algae, microorganisms, and macrophytes, sorption of dissolved organic or... American FWS wetlands was 2.2 mg/l, compared with an average inflow concentration of 4 .9 mg/l The average ammonium removal on a mass and percentage basis was 0.35 kg/ha-day and 38 percent, respectively Mean inflow and outflow values for ammonium for SSF wetlands were 6.0 and 4.5 mg/l, with mass and percentage removal rates of 0.62 kg/ha per day and 9 percent, respectively (Kadlec and Knight, 199 6) Biological . submerged macrophyte-dominated treatment wetlands (DeBusk et al., 199 0; Kadlec and Knight, 199 6) reveal that flow patterns depart widely from ideal plug-flow characteristics. Temperature-related water. a plug-flow reactor followed by multiple CSTRs (Kadlec, 199 7; King et al., 199 7). Recognition of nonideal flow characteristics, as documented by full-scale tracer studies, has led to most treatment wetlands. Thomas A. et al Wetlands for Water Treatment” Applied Wetlands Science and Technology Editor Donald M. Kent Boca Raton: CRC Press LLC,2001 ©2001 CRC Press LLC CHAPTER 9 Wetlands for Water