Performance studies of constructed wetlands for treatment of wastewater under temperate climatic conditions

Location of constructed wetlands site

Phytoremediation also termed as constructed wetlands is the natural method for the removal of impurities from wastewater. The aquatic vegetation is performing the task of wastewater treatment since the inception of the earth. Constructed wetlands have reportedly been used and found to be most efficient in the treatment of wastewater in various countries like the Czech Republic (Vyamazal 2010) China (Xiangyu et al., 2020), France (Morvannou A et al., 2015), Oman (Stefanakis, 2020), because of checking the efficacy of this treatment in a temperate zone like India, the efforts are being taken to compare the constructed wetland, a natural treatment by constructing a pilot plant and conventional activated sludge treatment. The pilot plant constructed wetland site is located on the premises of Uttarakhand Jal Nigam sewage treatment plant at Haridwar in north India. It is about 30 km from IIT Roorkee. This site was selected considering the following advantages:

The ample quantity of domestic wastewater (raw as well as treated) is available round the clock at the wetland site. A sufficient head was available to get both raw and treated wastewater by gravity at the site hence the cost of pumping is not required to bring it to constructed wetland site. As both the wastewater samples were available, the comparison of the performance of constructed wetlands could be possible with the performance of the activated sludge treatment process. Well-equipped laboratory facilities were available at the site to store and preserve the wastewater samples whenever needed.

Wastewater after preliminary treatment has been tapped at the overhead distribution channel and diverted to constructed wetland site through a 250 m long and 50 mm diameter underground PVC pipeline as shown in Fig. 1. The effluent after treatment with constructed wetland has been disposed-off in the existing outlet channel of the treatment plant premises.

Collection of Wastewater

In the town of Haridwar, wastewater is presently being collected through the network of a sewerage system. Open channels in town join the underground sub-main and finally the main sewer line. The main sewer line feeds their discharge to the wet well located in different areas of town after passing through the racks and screens. The wastewater is retained here for some time and pumped to the main pumping station at the scheduled time. Eleven such pumping stations are established in different parts of town. The location of the main pumping station is about four km away from the town. The wastewater again undergoes preliminary treatment through rack and screen after entering the premises of the main pumping station. A set of two pumps 80 HP each and another set of two pumps 40 HP each have been installed at the main pumping station. These pumps are operated in turns depending upon the inflow of wastewater. Some amount of this discharge is supplied directly to the sewage farms for irrigation whereas 24 MLD wastewater is pumped to the plant for treatment.

Existing Treatment Method

The detailed layout plan of the existing activated sludge process and the constructed wetland treatment pilot scale plant is shown in Fig. 1. Wastewater initially passes through preliminary treatment units. The preliminary treatment consists of a mechanical rack and screen where the floating and other discrete impurities like tobacco pouch, trash, silt, etc. are arrested and separated from wastewater. The wastewater then flows through the overhead distribution channel.

This channel performs two functions:

1. It has a partial flume that measures wastewater discharged by the treatment plant.

2. It distributes discharge uniformly over three units of the activated sludge process

The wastewater subsequently enters into the circular primary and secondary settling tanks with an aeration tank in between them. After treatment, wastewater enters into a collecting channel, from where it is disposed of in the river Ganga. The conventional activated sludge treatment plant was constructed in 1980 with a design capacity of 24 MLD. Each unit of the activated sludge process has a capacity of 8 MLD (totaling about 24 MLD by three units). However, presently because of an increase in population, wastewater generated in town is of the order of 35 MLD. Because of this increase in load, the proposal for augmentation of the plant is presently under consideration.

Comparison of Constructed Wetland with Conventional Treatment

There are various parameters based on which the constructed wetlands are compared with the other conventional wastewater treatment methods. The comparison of constructed wetlands with conventional treatment units has been done based on the following aspects.

Comparison concerning pollutant removal

Micro Flora and Fauna: - The degree of abatement achieved in constructed wetlands is much higher as the number of micro-organisms (more than 2000 types of bacteria, protozoa, yeasts, and more than ten thousand fungi and algae) present in the bed is several times more (Trivedy, 1998) than in the case of conventional treatment units.

Pathogen Removal: - It is found that the significant removal of the pathogen can be achieved in constructed wetlands as compared to conventional treatment processes. (Lambe and Whitman 1987; Trivedy 1998; Verma Mahesh et al., 2021; Hilley 1995)

Variation in the handling of pollutants: - Constructed wetlands can remove a variety of pollutants in comparison to the conventional treatment methods/units. This is because a large number of microorganisms are available for degradation.

Chemical Requirement: - Use of chemicals in constructed wetlands is rare whereas it is quite frequent in conventional treatments for pH adjustment, flocculation, etc.

Sulfur Removal: - Sulphates are degraded into elemental non-toxic Sulphur in constructed wetlands whereas in conventional treatment units these compounds are converted into toxic sulfides.

Nitrogen and phosphorus Removal: - Reduction of Nitrogen and Phosphorus is reported to be lower in horizontal flow type CW, but comparable in the case of vertical flow type CW and the conventional treatment units.

Comparison concerning cost

Cost of Construction: - Cost of construction of constructed wetlands (USEPA 1993) & (Vyamazal 2010) is reported to be much less because they are simple to construct require unskilled or semi-skilled labor and use locally available material. On the other hand, the cost of construction of conventional treatment is very high because of the involvement of factors like; huge civil work (earthwork, masonry, etc.), and the construction of facilities. It needs a large amount of expenditure to establish the process, fabrication/procurement/fixation of mechanical and electrical parts, equipment, establishing connectivity, etc.

Cost of Operation and Maintenance: - A few unskilled laborers can achieve the operation of constructed wetlands. Further, the requirement for electrical and mechanical energy is negligible. On the other hand, highly skilled laborers are required and a large amount of electrical and mechanical energy is required for operating the conventional treatment units. As the number of equipment is used in the conventional processes, their maintenance in terms of frequent repairs, and wear and tear also involve additional expenditure. Therefore, the overall cost of operation and maintenance in the case of conventional treatment is much more than constructed wetlands.

Ancillary Benefits

Although the main aim of constructed wetlands is to treat the wastewater, in addition to this, there are many more benefits of these wetlands to the public and wildlife habitats like (Knight 1997)

1. Animal diversity in surface flow treatment wetlands typically incorporates the full range of groups found in large wetlands, including microscopic invertebrates, macroinvertebrates, fishes, reptiles, amphibians, birds, and mammals.

2. Constructed wetlands have been utilized for direct human life support. Some of the large constructed wetlands are open for hunting

3. Wetland vegetation like common reeds after harvesting can be used as thatch for roofing

4. Some wetlands are attractive habitats for furbearers such as Onadilla zibethica, Myocastor, Coypus, etc.

5. The big wetlands are beneficial for recreational activities such as hiking, jogging, biking, and wildlife study.

6. Constructed wetlands in some cases are used as outdoor laboratories

Pollutant Removal Mechanisms

“The accomplishment of removal of pollutants in a constructed wetland (Sundaravadivel and Vigneswaran, 2001) is believed to happen in the following ways:

1. Uptake of pollutants directly by the plants

2. Plants, as well as substrate media, offer a large surface area for the proliferation of microorganisms that degrades pollutants.

3. Sedimentation of solids occurs because of the decreasing velocity of flow while flowing through the wetland

4. Large particles are removed through root and reed masses

5. Nutrients like nitrates and phosphates are adsorbed by soil and substrate

6. Detention time of wetland allows natural die-off of pathogens

7. UV radiation and excretion of antibiotics by plants to destroy pathogens

Interaction with wetland vegetation, the water column, and the wetland substrate results in pollutant removal in the wetland. Table 1 shows the type of pollutants and removal processes. These processes may be categorized as physical, chemical, and biological processes and described below:

Present study

Construction Details of Wetlands

Two beds of horizontal flow type constructed wetlands were constructed at the site as shown in Fig. 2A and 2B. One bed was used for studies on raw domestic wastewater and the other bed for studies on improving wastewater quality already treated by the activated sludge process.

A brick masonry manhole was constructed at the inlet end of the bed. It was provided with a constant head tank facilitating uniform flow in the bed. The control valve provided in the manhole regulates the flow. Wastewater was uniformly distributed by perforated pipe placed semi-circular PVC channel wherein V-notches were cut at 10 cm intervals.

The outlet chamber of brick masonry was constructed at the farther end of the bed. A flexible pipe provided at the end of the collection pipe helps in controlling the wastewater level in the bed. The wastewater after treatment was collected in a 75 mm diameter perforated pipe provided at the bottom of the bed at the outlet and taken into the outlet chamber. The treated wastewater is then disposed of in the main drain carrying treated wastewater from the existing sewage treatment plant. The Inlet and outlet zones of the bed were 50 cm wide and filled with gravel of size 20 to 80 mm. The construction of the pilot scale constructed wetland bed is also shown in photograph 1.

The soil texture available at the site was silty clay having a hydraulic conductivity of the order of 10^− 8 m/s. Although the hydraulic conductivity of the bed was very low, still as a factor of safety, two layers of polyethylene sheet having a thickness of 0.30 mm were laid at the bottom and sides of the bed.

The natural sand in the riverbed was not suitable for the wetland bed. Sieve analysis was carried out (Lambe and Whitman, 1987) for removing too fine and too coarse portions of natural sand. The effective size (d₁₀) of 0.8 mm and the Uniformity Coefficient (d₆₀/d₁₀) of 8.52 was calculated for the dressed sand. Filter media and gravel were then filled simultaneously in different layers in the bed. The coefficient of permeability (K_f) of filter media has been calculated from the effective size (d₁₀) in mm of the media by the relation [ K_f (m/s) = (d₁₀)² /100] given by Hazen (Terzaghi and Peck, 1956; Punmia, 1994) and found to be 6.4 x 10^− 3 m/s.

An exhaustive study of vegetation was carried out in and around the area. The existing bogs, marshes, and natural wetlands were surveyed for the study of macrophytes. The Phragmites karka was the most efficient and hardy plant and was used at the project site.

Monitoring Programme

The pilot-scale constructed wetlands located at Jagjeetpur, Haridwar had been planted with Phragmites australis microphytes one year before and is now fully grown up with a very high density of microphytes as shown in Fig. 2B and it has been observed that the wetland unit is now ready to be used for carrying out a series of studies one after the other. These studies and their findings are discussed in the subsequent studies given below:

1. Effect of Hydraulic Loading on Removal

Hydraulic loading is defined as the quantity of wastewater in liter per day applied to the unit area of the bed. It can also be represented in terms of cm/d. It gives an idea about the depth of water uniformly applied to the wetland surface over a specified time interval.

Mathematically, it is represented as

Hydraulic Loading Rate = Discharge (m³ / d)/ Plan Area (m²)

The reported range of hydraulic retention time for subsurface constructed wetlands (Crites, 1994) is between 2 to 7 days. While conducting the present study on the constructed wetland research site at Haridwar, the bed was set at the hydraulic loading corresponding to HRT ranging from 1 to 4 days at an interval of 1 day in decreasing order. It was practically impossible to set the discharge corresponding to HRT greater than four days because the hindrance caused by floating impurities at low flows resulted in frequent choking of the constant head tank outlet. After setting the bed for a particular hydraulic retention time, the bed was allowed to stabilize for one month in order to get uniform outlet BOD concentration. The inlet and outlet wastewater samples were then collected daily for the next one month and transported to the water quality laboratory for analysis.

While planning to observe the effect of discharge, hydraulic and organic loading on the percent removal of impurities in the wetland bed. The height of the outlet pipe in the outlet chamber has been adjusted and fixed by clamping so that the depth of wastewater inside the bed remains constant and 20 cm below the top surface of the gravel substrate of the wetland bed.

The inlet discharge valve provided in the inlet chamber is adjusted in such a way that the hydraulic retention time of wastewater in bed remains 4 days and associated values of discharge hydraulic loading, and organic loading 0.88 m³/d, 6 cm/day, and 62.5 kg/ha. d respectively is maintained throughout studies.

Similarly, for the HRT of 1 day, discharge, hydraulic loading, and organic loading were 3.5 m³/d, 24 cm/d, and 250 kg/ ha. d respectively. The remaining values fell in between the above range as shown in Fig. 3 and presented in Table 2.

The procedure is repeated for four different hydraulic retention times and associated discharge, hydraulic loading, and organic loading, and the respective percent removal of various parameters is given in Table No. 2

 

2. Estimation of BOD removal rate constant (K_t) of the treatment system.

It has been reportedly observed that the effluent BOD concentration in wastewater decreases with an increase in retention time. The data has been analyzed to check the variation in the ratio of effluent to influent BOD (i.e. C_e/C_o) with change in hydraulic retention time.

A first-order plug flow model for BOD removal has been reportedly used in earlier studies for the design of subsurface flow wetland systems. The general form of the model used in the United States, Europe, and Australia is presented below:

Ce/Co = e^{− Kt x.t}

Where,

C_e = effluent BOD (mg/l) C_o = influent BOD (mg/l)

K  _t  = rate constant (d^− 1)

t = hydraulic retention time (d)

The rate constant is temperature-dependent and follows van’t Hoff Arrhenius equation:

Kt = K₂₀ (θ)^T−20

Where

K₂₀ = BOD removal rate constant at 20 ⁰C (d^− 1)

T = Temperature of liquid in the system ⁰C

θ = Temperature activity coefficient

The temperature activity coefficient of 1.06 is indicated by EPA (USEPA, 1993; USEPA, 1988), hence the same value has been used for converting the decay rate constant from 24 ⁰C to 20 ⁰C in the present study.

The value of the rate constant (K₂₀) used in the above equation for the design of surface flow constructed wetlands has been reported to lie in the range of 0.8 to 1.104 d^− 1 (USEPA, 1988; WPCF, 1990; Crites, 1994). It is believed that; the rate constant for subsurface flow wetlands is higher than those for facultative lagoons or free water surface flow wetlands because of the availability of more surface area of the media. This surface area supports the development and retention of attached growth microorganisms, which are believed to provide most of the treatment responses in the system.

The coefficient (C_e/C_o) has been calculated for different hydraulic retention time values as presented in Table 3. A graph between C_e/C_o and hydraulic retention time has been plotted as shown in Fig. 4. Non-linear regression of the data yielded the following equation, which indicates the estimate of constant K_t at 24 ^oC as 1.0042 d^− 1.

Ce/Co ₌ e^{1.0042 x t}

The value of K_t at 20 ⁰C standard temperature has been calculated as:

K ₂₀ = 1.0042 / (1.06)²⁴⁻²⁰

= 0.795 d^− 1

The value of the removal rate constant (K₂₀) estimated as above for the present study is close to the range of 0.8 to 1.1 d^− 1 and is well supported by earlier studies (Crites, 1994; USEPA, 1988; WPCF, 1990; Hammer, 1989; Cooper, 1990; Bavor et al., 1988; Reed and Brown, 1992; Conley et al., 1991; Cooper and Hobson, 1990) (Crites, 1994; Cooper, 1990; Bavor et al., 1988; Reed and Brown, 1992; Conley et al., 1991; Cooper and Hobson, 1990).

Table 3: First Order BOD (n = 15, σ = ±15) Removal Rate Constant (K_t)

3. Determination of discharge, retention time, hydraulic loading, and organic loading at an effluent BOD  ₅  of 20 mg/l

As per the Bureau of Indian Standards (BIS, 1973) guidelines for disposal of wastewater in inland water bodies, the treated effluent should not have a BOD of more than 20 mg/l. It was observed from the experiments that the effluent BOD decreases with an increase in retention time. A graph between effluent BOD (C_e) and hydraulic retention time indicates the trend as shown in Fig. 5. To ensure that the pilot-constructed wetland unit operates by BIS norms, the HRT required to get the effluent BOD of 20 mg/l was determined from this figure. It has been observed that the hydraulic retention time of 1.75 days is sufficient for bringing the effluent BOD to 20 mg/l by the employed horizontal subsurface flow constructed wetland system. The respective discharge, hydraulic loading rate, and organic loading rate have been calculated as given below:

Discharge = Length x Width x Depth x Porosity/Hydraulic Retention Time

Discharge = (4.88 m x 3.1 m x 0.8 m x 0.29)/(1.75 d)

= 2.00 m³/d

Hydraulic Loading = Discharge/Area

Hydraulic Loading = 2000 (l /d)/4.88 (m) x 3.1(m)

= 132.21 l/m². d (13.22 cm/d)

Organic Loading = BOD X Discharge /Plan Area

Organic Loading = 108 x 2/4.88 (m) x 3.1(m)

= 14.28 g/ m². d (142.8 Kg/ha. d)

The estimated value of a desirable hydraulic retention time of 1.75 days is close to the range of 2–7 days recommended by various researchers. It has also been reported (USEPA, 1993) that several horizontal sub-surface flow constructed wetlands in the US (Mandeville 0.9 d, Monterery 0.9 d, Greenleaves Subdivision 1.0 d) are successfully operating at hydraulic retention time values close to the estimated value in this study

4. Comparison of removal efficiency of constructed wetland with activated sludge treatment.

The pilot-scale constructed wetlands site is located on the premises of the Uttaranchal Jal Nigam sewage treatment plant at Haridwar. The Jal Nigam is treating this domestic wastewater with an activated sludge process.

Samples were collected at a weekly interval from the inlet and outlet chambers of constructed wetland bed and outlet of the secondary settling tank of the existing activated sludge treatment for assessing the performance of both constructed wetland treatment and existing activated sludge treatment. The performance of both constructed wetlands and the activated sludge process has been assessed for three months and shown in Table No. 4 and Fig. 6