Performance of three type of constructed wetland systems for treating municipal waste water

A pilot constructed wetland systems project was constructed during 2015 at the University of Basrah, Iraq. These systems are a vertical subsurface flow system (VSSF), a horizontal subsurface flow system (HSSF) and a surface flow system (SF). These systems were planted with Phragmites australis, Typha domingensis and Certophyllum demersum respectively. It had been operated during 2016 as separated systems. The results recorded a total mean of 78.98% of NH 4 -N removal efficiency with 78.68% by VSSF, 76.04% by HSSF and 82.20 % by SF. This figure reached 90.58% removal of PO 4 , with 90.29 by VSSF, 90.18% by HSSF and 92.02 by SF. Also high level of total mean removal efficiency of 95.96% of BOD 5 , the results were 97.65% for VSSF, 97.99% for HSSF and 92.25% for SF. The results indicated that the system was highly effective at removing the target pollutants.


Introduction
Since first developed, constructed wetlands and their significant benefits have been extensively considered and widely utilized for treating wastewater from a variety of sources such as domestic, industrial and mine waste waters. 1,2 They have been effectively utilized for treating public wastewater. 3 Also, they have been commonly used to polish the final discharge of wastewater from treatment plants. 4 For many valid reasons, constructed wetlands could be the greatest mechanism for managing final wastewater treatment. 5,6 Constructed wetlands are now one of the most internationally diffused technologies for the biological, physical and chemical processes that occur in natural wetland. This paper focuses on a practical attempt to understand and implement some constructed wetland systems for treating waste water in Iraqi's weather conditions.

Materials and Methods
The constructed wetland station was contained within two parallel lines of three systems -vertical subsurface flow system (VSSF) planted with Phragmites australis, a horizontal subsurface flow system (HSSF) planted with Typha domingensis and surface flow system (SF) planted with Certophyllum demersum. Systems were made from fiber glass with the following dimensions: 300cm length, 120cm width and 100cm high. Also, three PVC lines were used to connect all the systems together Fig. 1 and Fig. 2. To test the systems separately, specific operation method was conducted, called "stable operation" whereby the systems operated during the first group of experiments during March, April and May. In this method, wastewater feeds into each system and remains for five days. Additionally, two levels of loading rate were used to test the system's volume ability for treating wastewater. These loading rates were 25% and 50% which equal to (162 and 324) L.

Results and Discussion
To evaluate the ability of a constructed wetland systems station in treating wastewater, some important parameters have been measured as the following: BOD5 with system type: The results in (Fig. 3) indicated that the total mean and Std. deviation of BOD5 for the different systems, namely: HSSF, SF and VSSF were 1.05±0.90, 4.19±3.72 and 1.23±0.94, and these values were 50.71±10.68 for SW and 2.40±0.32 1.23±0.94, for TW. The highest and lowest BOD5 levels were 11.33 and 0.10 which were recorded on day two at SF and on day three at HSSF respectively. Statistical analysis showed that there were no significant differences in BOD5 among retention time in day at P ≤ 0.05. The system type also showed no significant differences, the interaction between Retention Time * system type showed no significant differences at P ≤ 0.05.

Fig. 3: BOD5-with type of system in stable operation method
The removal efficiency of BOD5 after five days of treatment was 96.38%, while the maximum and minimum removal efficiency was 99.45% and 91.05% respectively. This value reached more than 90% after the second day of treatment. In terms of each system's ability to remove BOD5, the results were 97.99% for HSSF, 97.65% for VSSF and 92.25% for SF Fig. 4 and 5.  (Fig. 6) showed the total mean and Std. deviation of BOD5 with different loading rate percentages which were 1.74±1.74 for at a 25% loading rate and 2.68±3.47 for at 50% loading rate. These levels were 50.71±10.68 for SW and 2.40±0.32 for TW. The highest BOD5 value was 11.33, which was recorded within the SF system when the loading rate was 50%, while the lowest BOD5 level was 0.10 which was recorded within the HSSF system when the loading rate was 25%. Statistical analysis confirmed that there were no significant differences in BOD5 values between loading rate percentage at P≤0.05.

Fig. 6: BOD5-with the percentage of loading rate in stable operation method
The removal efficiency of BOD5 with both loading rate were 96.83% and 94.88% when the loading rates were 25% and 50% respectively (Fig. 7).

Fig. 7: BOD5 removal effciency with loading rate
All constructed wetland system showed an excellent capacity to reduce the BOD5 value to an accepted level in a very short time as clearly seen in Fig. 3. Generally, the statistical analysis indicated that there were no significant differences among single systems, which is in line with the findings of. 7 Wastewater with BOD5 above 300 mg/l is considered to be strong, while a BOD5 of less than 100 mg/l is considered weak. In order to prevent reduction of DO in water bodies, it is necessary to remove oxygen-demanding materials in influent water.
Organic matter could be breakdown by aerobic bacteria which works to utilze oxygen and produce biomass and energy. On the other hand, CH4 can be produced by anaerobic bacteria. 8 The results of this study achieved high removal efficiencies of 97.65%, 97.99% and 92.25% within the VSSF, VSSF and SF systems respectively as shown in Fig. 5. The results recorded a high level of BOD5 removal compared to many other previous studies. It was found by Olson et al. 9 that removal efficiency of BOD5 was about 87% in integrated system of septic tank flowed by a SF-CW system in Egypt. The average of 10.5-9.9mg/l was the final discharge of BOD5 after crossing planted beds in the Czech Republic's HSF-CW system, with an average removal efficiency of about 88%. 10 Also, BOD5 removal efficiency of 82% was achieved by 11 throughout an average of three years' treatment using a VSSF-CW system. An example from Pakistan also showed effective BOD5 removal of about 75% after five days of treatment using a Phragmaties constructed wetland bed. 12 The results give a clear picture showing the ability of these systems to significantly reduce the BOD5 level regardless of the high percentage of loading rate. In other words, even with a high loading rate of 50%, the results in figure 7 showed that the removal efficiencies were 96.83% when the loading rate was 25%, and this percentage was still very high sitting at about 94.88% when the loading rate was 50%. This means the amount of feeding wastewater can be increased within the system even if it is up to the normal ability of these system for SW. This evidence also could provide a good indicator of constructed wetland systems' propensity for removing BOD5, especially in our environmental circumstances which can reduce the area required to design the constructed wetland system. For example, in China, a study compared seasonal variation of removal efficiency between cold and warm temperatures, where the results showed the following: 92%, 73% and 71% for COD, BOD5 and NH4-N at warmer temperatures, while it dropped to 85%, 40% and 20% during cold weather. 13 NH4-N with system type: The results in (Fig. 8) illustrate a dramatic drop in the value of NH4-N. The total mean and Std. deviation for each system was 4.20 ± 2.89 for HSSF, 3.54±2.78 for SF and 3.91 ± 2.37 for VSSF. These values were about 19.60±7.41 for SW and 8.40±5.94 for TW. Statistical analysis showed that there were significant differences in NH4-N values among retention time in days at P≤0.05. There were no significant differences in NH4-N among system type. In addition to that, the interaction between Retention Time * system type showed no significant differences at P ≤ 0.05.

Fig. 8: NH4-N with the type of system in stable operation days
Removal efficiency of NH4-N was achieved at about 90.14% after five day of treatment, whereas 87.50% was removed after the second day of treatment. Also the total mean of removal efficiency was 76.04%, 82.20% and 78.68% for HSSF, SF and VSSF respectively ( Fig. 9 and  10).  (Fig. 11) represented the relationship between the loading flow rate percentage and removal of NH4-N. It shows that there was a high removal percentage of amonium when the loading flow rate was 25%, and 50%. The total mean and Std. deviation were 3.92±2.67 mg/l when the loading rate percentage was 25%, and 2.80 mg/l when the loading rate percentage was 50%. The maximum value was 28 mg/l for SW, while the minimum was 1.40 mg/l which was recorded at HSSF, SF and VSSF when the loading rate percentage was 25%. Statistical analysis confirmed that there were no significant differences in NH4-N between loading rate percentage at P ≤ 0.05.

Fig. 11: NH4 with percentage of loading rate in stable operation method
The removal efficiency for both loading rates, which were 25% and 50%, were 77.22% and 98.33% (Fig. 12).   9 Valipour and Ahn, 2016 pointed out that microorganisms continue to grow and utlize organic matter and nitorgen with expanded of hydraulic retention time. The system's performance recorded a high removal efficiency within the SF system with about 82.20% flowed by VSSF with about 78.68%. The lowest removal efficiency was recorded within the HSSF system. As a result of its limited capacity to transfer oxygen, HSSF has less ability to oxidize NH3-N to NO3-N. 14 In addition, the VSSF system is well-known for its ability to transfer oxygen due to intermittent feeding which adds advantage to this type of system and increases its removal efficiency. 15 It has been reported that the SF constructed wetland systems have had high removal efficiency of organic compounds through settling and biological degredation. Also, some important processes such as nitrification, denitrification and ammonia volatilization take place especially under alkline circumstances. This high pH is a result of algal photosynthesis decay. 16 The same trend of NH3-N removal efficiency results (about 96%) have been achieved through using a gravelbased hybrid system constructed wetland to treat wastewater. 17 Also, the performance of Will's Barn vertical flow constructed wetland recorded a high average decline of NH4-N from about 93.9 mg/l to 10.29 mg/l in effluent treated water. 18 In addition, high removal of TN-N and NH4-N was observed in the vertical flow system, whereas horizontal flow showed a high removal efficiency of COD compared with VSSF. 19 Moreover, after crosing the two stages of VSSF-CW, NH4-N sharply decreased from 38mg/l to 7.3mg/l after the first stage and to about 2mg/l after the second stage of the constructed wetland. 20 Aorthou-phosphate with system type: The results in (Fig. 13) illustrated that orthou-phosphate was declined in all system but in different levels. The total mean and Std. deviation for HSSF, SF and VSSF were as respect 2.39±2.84, 2.02±3.98 and 2.42± 4.27 mg/l respectively. These values were 16.16±14.14 and 1.01±1.23 mg/l for SW and TW respectively. The maximum and minimum values were 26.86 and 0.03 mg/l for SW and VSSF during fifth day respectively. Statistical analysis confirmed that there were significant differences in orthou-PO4 between retention time in days at P ≤ 0.05. However, there were no significant differences in orthou-PO4 among system type at the same level. The interaction between system type and months also showed no significant differences in orthou-PO4 values.

Fig. 13: PO4-with the type of system in stable operation method
Removal efficiency of PO4 after five days of treatment reached 92.56%. Systems preformance of removal efficiency recorded high achievement in Sf with high percentage of 92.02 % followed by 90.29% and 90.18% for VSSF and HSSF respectively ( Fig. 14 and  15). Orthou-phosphate with loading rate percentage: The results in (Fig. 16) showed that all systems removed approximately the same amount of PO4 during the feeding systems with 25% loading flow rate of sewage water. Also, SF removed high amount of PO4 even the loading rate was 50% other systems (including HSSF and VSSF) were not tasted with 50% of loading rate. The total mean and Std. deviation of ortho-PO4 for both loading rate percentage 25% and 50% were 2.38±3.82 and 0.87±0.30 mg/l respectively. The highest and lowest points recorded were 26.86 and 0.03 mg/l which were measured in SW and VSSF when the loading rate was 25%. Statistical analysis showed that there were no significant differences in orthou-PO4 between loading rate percentage at P ≤ 0.05.

Fig. 16: PO4-with the percentage of loading rate in stable operation method
High removal efficiency of PO4 were achieved with both loading rate as the removal efficiency was 90.48% when the loading rate was 25 % and the results were much better when the loading rate was 50% as it reached about 95.87%; however, this result include measuring of PO4 for SF system only (Fig. 17).

Fig. 17: PO4 removal efficiency with loading rate in stable operation
Results of this study achieved excellent results compared to many previous studies conducted in order to evaluate PO4 removal efficiency of constructed wetlands (Fig. 13 & 14). While sewage water had a total mean of about 16mg/l of PO4, this amount is removed to about 0.92, 1.51 and 3.56 in the SF, VSSF and HSSF systems respectively by the fifth day of experimentation. However, a clear decrease of PO4 was recorded after the first and second days of treatment. Also, it has been noticed that the systems had a similar ability to remove PO4 with the highest removal percentage achieved by SF, followed by the VSSF and HSSF systems respectively.
PO4 is removed within a constructed wetland system by several mechanisms including adsorbtion, precipitation and plant uptake, which is considered temporal storage as the nutrients could be released to the aquatic environment after the plants have decayed. 16 Phosphorus removal efficiency within constructed wetlands, as reported in many previous studies, significantly varied from 6-99%, dependent primarily on wetland design, loading rate and environmental condition. 21 As an example, the reduction of phophorus was about 66% in a lab-scale of constructed wetland. 22 The results of removing PO4 within the SF system with that level of loading rate was higher compared with a 25% loading rate. As shown in figure (17), removal efficiency reached around 95.87% with a loading rate of 50%, whereas the removal efficiency of PO4 was 90.48% when the loading rate was 25%. A possible reason for this is the high growth of plant Certophyllum demersum within the SF system which could take more PO4 for its growth.

Conclusion
Overall it can be clearly indicated that, implementation of constructed wetland systems could be a valid solution for treating wastewater as all systems (VSSF, HSSF and SF) which have been implemented at this experiment showed an excellent result in order to remove a high percentage of the target pollutants.