Vegetated beds

Kickuth's concept was closer to the traditional under standing of a soil treatment of sewage but his statement that root and rhizome growth would improve hydraulic conductivity of heavy soils failed on several early con struction sites and it gave a harsh setback to the scientific and official acknowledgment of HF CWs as sewage treatment systems. The problem in constructed beds where soil is used as medium is the low hydraulic conductivity which results in surface flow and thereby a short retention time within the system. HSSF CWs were introduced to Denmark in 1983 and then to the United Kingdom in 1985. The operational experience from more than 100 HSSF systems in these two coun tries resulted in the use of coarse materials such as gravel or crushed rock with fraction in the range of 5-20 mm (Figure 5).

Figure 3 Horizontal grit chamber at HSSF CW at Kamen, Czech Republic. Photo by J. Vymazal.

The following equation, first proposed by Kickuth, has been widely used for sizing of HSSF systems for domestic sewage treatment:

Ah = n Cin ln Cout) /Kbod where Ah = surface flow of bed (m2), = average flow (m3 d J) Cin = influent BOD5 (mgl J), Cout = effluent BOD5 (mgl J), and KBoD = rate constant (m d J).

There were a lot of discussions on the KBOD value. Formerly proposed value of 0.19 m d 1 by Kickuth resulted in too small area of the bed (about 2 m2 per one PE, population equivalent) and consequently lower treatment effect. The field measurements in operational systems showed that the value of KBoD is usually lower (0.070.1 m d 1). It has also been found that rate constants gen erally increase with hydraulic loading rate and BOD5 mass loading rate. In addition, in systems which have been in operation for at least 10 years a steady increase in KBoD value was observed. At present, using field measurements from many operational systems, the value of 0.1 m d 1 is considered as sufficient but in many countries values around 0.08 m d 1 are used. This generally means that the value of Ah is about 5 m2 PE 1. However, this sizing is suitable for BOD5 and suspended solids removal but not appropriate for removal of nitrogen and phosphorus. The

Figure 6 Vegetated bed configuration of HSSF CWs. From Vymazal J, Brix H, Cooper PF, Green MB, and Haberl R (eds.) (1998) Constructed Wetlands for Wastewater Treatment in Europe. Leiden, The Netherlands: Backhuys Publishers.

Figure 5 Example of a large single bed (100 x 35 m) -Freethorpe, United Kingdom. Photo by J. Vymazal.

data from 66 Danish HSSF CWs revealed K values for total N (0.033 md *) and total P (0.025 md *). In North America, for size calculations, background values of indivi dual pollutants are often used (see Constructed Wetlands, Surface Flow).

The early HSSF CWs frequently used only one bed even for large areas (Figure 5). It created many problems with maintenance of the system and especially vegetation.

Figure 6 Vegetated bed configuration of HSSF CWs. From Vymazal J, Brix H, Cooper PF, Green MB, and Haberl R (eds.) (1998) Constructed Wetlands for Wastewater Treatment in Europe. Leiden, The Netherlands: Backhuys Publishers.

Also, it was difficult to distribute wastewater evenly and avoid short circuiting. At present, the area of larger CWs is usually divided into several beds in order to provide better access for maintenance and better wastewater dis tribution (see Figure 2). On the other hand, this approach requires more land. For example, CW for tertiary treat ment of 24435 PE in Beja, Portugal, consists of 32 beds with a total area of 22 800 m2. Examples of various bed configurations are given in Figure 6. The aspect ratio (length:width) also changed during years. While the early systems used to have long and narrow trenches with aspect ratio (length:width) >1, aspect ratio is usually <1, at present. Wide inlet zone prevents overloading but on the other hand it is more difficult to maintain an even distribution across the inlet zone.

The depth of vegetated beds with HSSF was initially based on the requirement that roots and rhizomes of the vegetation should penetrate the full depth of the bed in order to eliminate totally anaerobic zones. As the roots of the most frequently used plant, common reed (P. australis), are capable to successfully penetrate to the depth of about 0.6 m and start to weaken beyond that point, the recommended bed depth was 0.6 m. Roots and rhizomes of reeds and all other wetland plants are hollow and contain air filled channels that are connected to the atmosphere for the purpose of transporting oxygen to the root system. The majority of this oxygen is used by the roots and rhizomes themselves for respiration, but as the roots are not completely gas tight, some oxygen is lost to the rhizosphere. Although it has been proven that oxygen transported from aboveground organs diffuses only to the thin substrate layer adjacent to the roots and rhizomes the typical arrangement of HSSF CW has the depth of filtration bed usually 0.6-0.8 m. According to the working principle of HSSF CWs, the amount of oxygen released from roots and rhizomes should be sufficient to meet the demand for aerobic degradation of oxygen consum ing substances in the wastewater as well as for nitrification of the ammonia. However, many studies have shown that the oxygen release from roots of different macrophytes is far less than the amount needed for aerobic degradation of the oxy gen consuming substances delivered with sewage.

A wide variety of materials is used as a substrate for HSSF CWs. The early systems used cohesive soils with high clay content. However, the low hydraulic conduc tivity of such materials resulted in clogging problems. Since the late 1980s, materials with high hydraulic con ductivity have been used, namely gravel (pea, cobble) or crushed rock (Figure 7) with the size fractions in the range of 5-20 mm. It is recommended to use one fraction and do not mix several fractions together as this may result in uneven homogeneity of the filtration material. The coarse material provides high treatment effect, does not clog, and supports good plant growth. It is also

Figure 7 Pea gravel at newly built HSSF CW Sedlce, Czech Republic. Photo by J. Vymazal.

recommended to use washed materials because this removes fines which could block the void spaces.

CWs with HSSF are usually sealed in order to prevent uncontrolled water seepage into groundwater. Most of the systems have used a plastic liner or membranes such as HDPE, LDPE, or PVC 0.5-2 mm thick (Figure 8). Where local subsoil has low hydraulic conductivity (approx. 10 8 ms 1 or less), it is not necessary to use plastic liners. In order to prevent liner damage by filtration material particles geotextiles are commonly used to cover plastic liner. Also, geotextiles could be also used beneath the liner.

Distribution and collection of wastewater

In HSSF CWs the aim is to get even distribution across the full cross sectional area of the inlet end of the bed. The early systems used either open trenches (Figure 9) or channels with castelated or vee notch weirs (Figure 10). However, they had problems with screenable material col lecting at edges and causing maldistribution. These were also very expensive to construct and so there was a ten dency to move to single pipes. Distribution pipes are either buried under the surface of distribution zone (Figure 11) or laid on the surface (Figure 12) and overlayered with stones

Figure 8 Plastic liner used to seal the bed at HSSF CW Rudikov, Czech Republic. Photo by J. Vymazal.

Figure 7 Pea gravel at newly built HSSF CW Sedlce, Czech Republic. Photo by J. Vymazal.

Figure 8 Plastic liner used to seal the bed at HSSF CW Rudikov, Czech Republic. Photo by J. Vymazal.

Figure 9 Inflow open trench in HSSF CW Granfeld, Denmark. Photo by J. Vymazal.
Figure 10 Distribution channel at HSSF CW at Othfresen, Germany. Photo courtesy Hans Brix.
Figure 11 Distribution pipes buried under the surface of the distribution zone. In this case distribution zone and vegetated beds are at the same level. In front of the picture there is a pipe for distribution system cleaning (HSSF CW Sedlce, Czech Republic). Photo by J. Vymazal.

Figure 12 Distribution pipe on the surface of the distribution zone in HSSF CWs at DraZovice, Czech Republic. Photo by J. Vymazal.

Figure 13 Distribution zone in HSSF CW Pribraz, Czech Republic. Distribution pipe is positioned at the same level as filtration bed surface and covered by approx. 0.5 m layer of large stones. Photo by J. Vymazal.

(Figure 13). It is essentail that the distributor be cleaned out by brushing and/or flushing (Figure 11). Distribution zones, 0.5-1.0 m wide, are filled with stones in the size range of 50-200 mm. The large stones are necessary for good water distribution across the inlet zone.

The outlet collector pipes are made of the same materi al and use the same diameters as distribution pipes and are buried near the bottom of the outlet zone (see Figure 1). Outlet zone, usually 1-2 m wide, is filled with the same material as distribution zones. The outlet pipe is con nected through the liner to the outlet pump (Figure 1 ). Several structures for water level control are used. The early systems mostly used a 90o elbow which allows to set the water level in the bed by rotating the elbow sideways. However, the steel elbows suffered from rusting and were

difficult to operate. Much better performance is provided by the use of plastic elbows (Figure 14) or flexible hoses (Figure 15) or combination of these two options (Figure 16). This arrangement provides very fine and easy control of water level. The water level is usually maintained 5-15 cm below the bed surface.


The macrophytes growing in CWs have several proper ties in relation to the treatment processes that make them an essential component of the design. The most important effects of the macrophytes in HSSF CWs in relation to wastewater treatment processes are the physical effects the plant tissues give rise to, such as insulation of the bed surface during the winter or provision of surface area for attached microorganisms in the filtration bed. The meta bolism of the macrophytes (plant uptake, oxygen release) are of minor importance in HSSF CWs. The macrophytes have other site specific valuable functions, such as pro viding a suitable habitat for wildlife, and giving systems an esthetic appearance. It has been shown that in tempe rate and cold climate regions, probably the most important role of plants is to insulate the bed during winter. Therefore, it is desirable to use fast growing spe cies which also have a high aboveground as well as belowground biomass. The macrophytes for HSSF CWs should also be tolerant of a relatively high organic load in the wastewater.

These characteristics are found in helophytes, that is, emergent vascular plants growing in marshes. Among them, P. australis (Common reed, Figure 17), is the most widely used plant in Europe, Australia, and Asia. In some countries, Phragmites is used exclusively - for example, in the United Kingdom, where HSSF CWs are called 'Reed bed treatment systems'. In Europe, Phalaris arundinacea (Reed canarygrass, Figure 18), Glyceria maxima (Sweet

Figure 18 Phalaris arundinacea (Reed canarygrass). Photo by J. Vymazal.
Figure 19 Iris pseudacorus (Yellow flag). Photo by J. Vymazal.

mannagrass), or various Cattails (Typha spp.) are used as well. Plants used for on site treatment systems are usually decorative species such as Iris pseudacorus (Yellow flag, Figure 19), Iris sibirica (Blue flag), Filipéndula ulmaria (Green of the meadow), Epilobium hir sutum (Hairy willow herb), or Lythrum salicaria (Purple loosestrife). In many areas of the Unites States, Phragmites is considered an exotic and invasive species by natural resource agencies and as a result, use of this species in North America has been limited. Examples of plants used in North America are Colocasia esculenta (Taro), Phalaris arundinacea, Iris versicolor (Northern blue flag), Saggitaria latifolia (Duck potato), or Scirpus fluviatilis (River bulrush). In Africa, Cyperus papryrus (Papyrus) and P. mauritianus are most frequently used for HSSF CWs.

Plant stock that are frequently used to establish CWs for wastewater treatment include seeds, bare root seedlings, rhizomes, greenhouse grown potted plants, and field harvested plants. It seems that bare root seedlings are the best option and plant stands establish very soon; also rhi zomes provide good results. Potted plants are not too suitable for HSSF CWs as the soil cloud will be washed out to the deeper layers and fill void spaces. Also, potted plants are not easy to transport as compared to bare root plants of rhizomes. The use of field harvested plants is usually cheap but very often introduces many weeds.

Weeds (plants that were not intentionally planted) occur mostly within vegetated bed margins. The most frequently occurring species are those that tolerate pollu tion and wet conditions as well, for example, Rumex obtusifolius (bitter dock), Urtica dioica (slimsting nettle), or Atriplex patula (fat hen saltweed). If high quality seedlings are used, planting is made carefully in a sufficient density (4-8 seedlings perm2) and in proper time of the year; weedy species are completely overgrown and their scatter growth is limited to margins. Plants in HSSF CWs are usually not harvested and the litter is left on the surface. However, plants could be harvested but it has been shown that plant harvesting removed only negligible amount of nutrients from wastewater. Also, if plants are harvested the systems look more esthetic as compared to systems which are not harvested. In developing countries the biomass is always used for manufacturing various pro ducts as mats, fences, roofs, etc.

Removal mechanisms

Aerobic degradation of soluble organic matter is governed by the aerobic heterotrophic bacteria. However, dissolved oxygen supply is very limited in HSSF CWs and, there fore, anoxic and anaerobic processes usually prevail. Anaerobic respiration occurs in the soil zone below the Fe3+ reduction zone and the process can be carried out by either facultative or obligate anaerobes. It represents one of the major ways in which high molecular weight carbo hydrates are broken down to low molecular weight organic compounds, usually as dissolved organic carbon, which are, in turn, available to microbes. Anaerobic degradation is a multistep process. In the first step the primary end products of fermentation are fatty acids such as acetic, butyric, and lactic acids, alcohols, and the gases CO2 and H2. Strictly anaerobic sulfate reducing bacteria and methane forming bacteria then utilize the end products of fermentation and, in fact, depend on the complex community of fermentative bacteria to supply substrate for their metabolic activities.

Suspended solids that are not removed in a pretreat ment system are effectively removed by filtration and settlement. Most of the suspended solids are filtered out and settled within the first few meters beyond the inlet zone. The accumulation of trapped solids is a major threat for good performance of HSSF systems as the solids may clog the bed. Therefore, the effective pretreatment is necessary for HSSF systems.

The major removal mechanism of nitrogen in HSSF CWs is nitrification/denitrification. Field measurements have shown that the oxygenation of the rhizosphere of HSSF CWs is insufficient and, therefore, incomplete nitrification (i.e., oxidation of ammonia to nitrate) is the major cause of limited nitrogen removal. In general, nitri fication which is performed by strictly aerobic bacteria is mostly restricted to areas adjacent to roots and rhizomes where oxygen leaks to the filtration media. On the other hand, prevailing anoxic and anaerobic conditions offer suitable conditions for denitrification but the supply of nitrate is limited as the major portion of nitrogen in sewage is in the form of ammonia. Volatilization, adsorp tion, and plant uptake play much less important role in nitrogen removal in HSSF CWs. Volatilization is not effective as there is no free water surface and adsorption is greatly limited by the fact that filtration media (gravel, rock) do not provide suitable sorption sites. Plant harvest ing contributes to an overall nitrogen removal only marginally (usually <10% of the inflow load) with N standing aboveground stocks in the range 20-60 gN m 2

Phosphorus is removed primarily by ligand exchange reactions, where phosphate displaces water or hydroxyls from the surface of Fe and Al hydrous oxides. However, media used for HSF wetlands (e.g., pea gravel, crushed stones) usually do not contain great quantities of Fe, Al, or

Ca and therefore, removal of phosphorus is generally low. Removal of phosphorus could be enhanced by the use of filtration media with high sorption capacity such as calcite or light weight clay aggregates (Leca). In such cases, how ever, the sorption capacity is always saturable and therefore, the media must be replaced in order to maintain high P removal. Removal via harvesting accounts usually for <5% of the inflow load with the P standing stock in the aboveground biomass in the range of 3-10gPgm . Harvesting may play a more significant role in nutrient removal in lightly loaded tertiary treatment systems.

The removal of microbiological pollution is very sel dom the primary target for constructed treatment wetlands. However, wetlands are known to act as excellent biofilters through a complex of physical, chemical, and biological factors which all participate in the reduction of the number of bacteria of anthropogenic origin. Physical factors include mechanical filtration and sedimentation; chemical factors include oxidation, UV radiation, expo sure to biocides excreted by some plants and adsorption to organic matter. Biological removal factors include antibio sis, predation by nematodes, protozoa, and zooplankton, attack by lytic bacteria and viruses and natural die off.

Treatment efficiency

HSSF CWs are very effective in removing organics (BOD5, COD), suspended solids, and microbial pollution (Tables 1 and 2). Removal of nutrients is lower but it is important to realize that most HSSF CWs have been designed with organics and suspended solids as primary target and therefore, the dimensioning is not sufficient for removal of nutrients (see section titled 'Vegetated beds'). However, the results in Table 1 present only the results of vegetated beds. When overall treatment effect is eval uated it is necessary to include also the treatment effect of pretreatment units. For nutrients, it usually does not represent any substantial removal but removal of sus pended solids, BOD5, and COD in pretreatment units

Table 1 Treatment efficiency of vegetated beds of HF CWs - worldwide experience

Inflow Outflow Efficiency

BOD5 108 16.0 85 164

COD 284 72 75 131

TSS 107 18.1 83 158

Data from Australia, Austria, Brazil, Canada, Czech Republic, Denmark, Germany, India, Mexico, New Zealand, Poland, Slovenia, Sweden, USA, and UK. FC, fecal coliforms.

From Vymazal J (2005) Constructed wetlands with horizontal subsurface flow and hybrid systems for wastewater treatment. Ecological Engineering 25: 478 490.

Table 2 Loading of constructed wetlands with horizontal subsurface flow





Efficiency (%)












































aInflow for BOD5, COD, and TSS is expressed in kg ha 1 day 1 and for TP, TN, NHJ N, and NO3 N in g m 2yr 1. bOutflow for BOD5, COD, and TSS is expressed in kg ha 1 day 1 and for TP, TN, NH| N, and NO3 Ningm 2yr 1. For details see Table 1.

From Vymazal J (2005) Constructed wetlands with horizontal subsurface flow and hybrid systems for wastewater treatment. Ecological Engineering 25: 478 490.

aInflow for BOD5, COD, and TSS is expressed in kg ha 1 day 1 and for TP, TN, NHJ N, and NO3 N in g m 2yr 1. bOutflow for BOD5, COD, and TSS is expressed in kg ha 1 day 1 and for TP, TN, NH| N, and NO3 Ningm 2yr 1. For details see Table 1.

From Vymazal J (2005) Constructed wetlands with horizontal subsurface flow and hybrid systems for wastewater treatment. Ecological Engineering 25: 478 490.








Coal, metals



Food processing

Heavy industry

Winery, cheese, potato, bakery, sugar, abattoir, fish, meat, milk

Winery, cheese, potato, bakery, sugar, abattoir, fish, meat, milk

Oil refineries, fertilizers, herbicides, explosives, chemicals, pulp and paper, metal finishing


Urban, highway, parking lots, airports, field crop, nursery, greenhouses, pastures

Landfill leachate

Figure 20 Types of wastewater treated in HSSF constructed wetlands.

usually amounts to 40-60%, 30-50%, and 20-40%, respectively, as compared to raw wastewater inflow.

CWs with HSSF are used for wide variety of waste waters (Figure 20). Probably, the most CWs have been built to treat municipal or domestic sewage. However, recently the HSSF CWs have been used for many other types of wastewater including those from agriculture, industry (Figure 21), or landfill leachate (Figure 22).

0 0

Post a comment