Optimization of constructed wetlands on purifying black-odorous water and their potential purification mechanism

Black-odorous water refers to the water bodies with unpleasant colors (often black) and terrible smells in urban regions. The black-odorous phenomenon is a serious water contaminant problem caused by urbanization, which has been reported worldwide, particularly in developing countries (Rixen et al. 2008; Le et al. 2010; Handa & Jadhav 2016). Their occurrences seriously affect the living condition of local residents, and poses a threat on human health and the ecosystem (Zhang et al. 2021). However, such wastewater is difficult reduce by municipal sewage treatment plants, due to its extensive pollutant sources and vast contaminated areas (Cao et al. 2020). Here, constructed wetlands (CWs) are regarded as the most desirable sustainable, low-cost technology providing physical, chemical, and biological processes for the primary purpose of water quality improvement (Imfeld et al. 2009; Vymazal 2019).

The purification mechanism of CWs is impacted by various factors, i.e. flow type, retention time, pH value, temperature, plant, and substrate (Vymazal 2007). Each pollutant has its specific removal pathway that is influenced by different factors. CWs own a strong ability on the removal of organic matters: the suspended organics are mainly eliminated via physicochemical processes, such as filtration and precipitation, the dissolved organics are primarily adsorbed to plant roots and filter biofilms, and subsequently degraded by microorganisms. There are many factors impacting the removal of nitrogen: the internal factors mostly contain flow types, filters, plants, and microbes, while the external factors consist of hydraulic retention time, oxygens, pH values, and carbon sources (Prochaska et al. 2007; Stefanakis & Tsihrintzis 2012). Although phosphorous is mainly removed through physical procedures in CWs, it is also impacted by microbial activities, plants, and hydraulic loading rates (HLRs) (Akratos & Tsihrintzis 2007). Hence, the configurations and operation parameter of CWs may decide the decontamination results. It is very necessary to assess the influence degree of each factor for achieving the optimal performance on the purification of black-odorous water.

In this study, nine CWs with different flow types, filters, plants, and HLRs were prepared to evaluate their capacity on the purification of black-odorous water in two seasons (summer vs winter). Our aims are (1) to optimize the key parameters of CWs for purifying black-odorous water, and (2) to give a better understanding for their potential purification mechanisms. The knowledge achieved from the present work will be useful for the application of CWs on the purification of black-odorous waterbodies.

After a stabilization period of half a year, water samples were collected continuously in summer (August, n = 6) and winter (February, n = 6) for evaluating their water quality parameters: (1) physio-chemical parameters, i.e. water temperature (WT), pH, dissolved oxygen (DO), and oxidation-reduction potential (ORP), were measured using an YSI water quality meter (YSI ProPlus, USA), (2) macro-pollutants parameters, i.e. total suspended solid (TSS), chemical oxygen demand (COD), ammonium (NH₄-N), nitrate (NO₃-N), total nitrogen (TN), phosphate (PO₄-P) and total phosphorus (TP), were determined following the standard methods (SEPA 2002).

Physio-chemical parameters, i.e. WT, pH, DO, and ORP, in the CWs are presented in Table 2. In spite of WT being higher in summer (28.8–29.3 °C) than in winter (12.0–16.0 °C), the pH, DO, and ORP values were comparable in both seasons, with neutral (6.97–7.17), anaerobic (0.56–1.50 mg/L), and reductive (−242 to −221 mV) conditions, respectively. In winter, the DO concentration was relatively low and increased from 0.56 mg/L of influent into 1.01 mg/L of effluent, indicating good oxygen transport capacity of CWs, particularly in VF-CWs (1.78–2.60 mg/L) (Stefanakis et al. 2014).

Removal efficiencies of macro-pollutants, i.e. TSS, COD, NH₄-N, NO₃-N, TN, PO₄-P, and TP, in the CWs are also depicted in Table 2. Among them, COD, PO₄-P, and TP were removed more efficiently in summer (72% of COD, 33% of PO₄-P, and 31% of TP) than in winter (50% of COD, 13% of PO₄-P, and 0% of TP). No significant difference was found in the removals of TSS, NH₄-N, NO₃-N, and TN between two seasons. During the whole experimental period, all the CWs performed well on the TSS removal (75–79%) and efficiently on the NH₄-N removal (39–53%). Collectively, the purification result of black-odorous in CWs was better in summer than in winter, possibly due to warm weather favors endothermic hydrolysis and biodegradation process (Truu et al. 2009). Meanwhile, some pollutants were removed equally in both seasons, which may be because cold climate is conductive to exothermic adsorption and therefore make up the lack of microbial effect.

Correlations between the value of physio-chemical parameters and the removal loading of macro-pollutant parameters in the CWs are exhibited in Table 3.

Significant positive relationships were found between WT and TSS/COD/ PO₄-P in winter, DO and ORP in both seasons, TSS and COD/TP in both seasons, COD and TP in winter, NH₄-N and TN in both seasons, and PO₄-P and TP in summer (Table 3). It can be seen that the organic matter, TN, and TP from black-odorous water mainly existed in the form of suspended organic, NH₄-N, and PO₄-P. In warm summer, the microbial activity was already high and was not obvious to be impacted by WT, however, in cold winter, WT became critical to the growth of microorganisms, thus affecting the removals of contaminants. Furthermore, the values of DO and ORP were interdependent to each other, which determined the distribution of dominant microorganisms (Liu et al. 2018; Li et al. 2019).

Correspondingly, significant negative relationships were observed between DO/ORP and NO₃-N in both seasons, and NO₃-N and PO₄-P in winter (Table 3). The lower the DO and ORP values are, the higher the NO₃-N removals are. Moreover, the removal efficiency of NO₃-N was very poor in CWs (<0%), whereas NH₄-N was reduced efficiently from black-odorous water (39–53%) (Table 2). This suggests that the generation of NO₃-N is much faster than its reduction. Hence, we believe that higher DO and ORP values are beneficial to aerobic nitrification in CWs, thus promoting the conversion from NH₄-N into NO₃-N (Wu et al. 2017).

Variances among different levels of factors according to orthogonal assay are listed in Table 4. Water quality parameters (physio-chemical and macro-pollutant parameters) were markedly impacted by flow type for DO and ORP in both seasons, for NO₃-N in summer, and for PO₄-P in winter. The TN removal was notably impacted by different filters in winter. Moreover, prominent influence was also observed in TSS and COD in both seasons, and in WT in winter.

Table 4 shows that flow type owns the greatest effect on physio-chemical parameters (i.e. DO and ORP) from black-odorous water, and this effect is much stronger in summer. For one thing, it demonstrated that the flow type could determine physio-chemical parameters such as ORP in CWs (Hijosa-Valsero et al. 2010), for another, such change might contribute to the distribution of specific microbial biomass in CWs, especially in warm conditions. Furthermore, the influence of flow type on the microenvironment of CWs is also reflected in the NO₃-N removal, since a significant positive correlation was obtained between the NO3-N removal and the DO and ORP value (Table 3).

Table 4 also displays that HLR plays an important role on the removal of macro-pollutants (i.e. TSS and COD). This is because HLR may decide the retention time of contaminants in CWs, and therefore change the adsorption ability of pollutants to filters and their reaction time with biofilms.

The DO and ORP values were higher in VF-CWs (1.78–3.26 mg/L and −189 to −164 mV) than in HF- and SF-CWs (0.35–0.96 mg/L and −258 to −236 mV). A similar phenomenon also appeared in the removal loading of PO₄-P (0.11–0.47 g/m²/d in VF-CWs and 0–0.32 g/m²/d in HF- and SF-CWs) (Figure 2). Since there was no significant relationship between the PO₄-P removal and the DO and ORP value (Table 3), it might be explained by the unsaturated system of VF-CWs that favor the adsorption process of PO₄-P on filters and their biofilms. Conversely, NO₃-N was removed more efficiently in HF- and SF-CWs (−0.03 to 0.19 g/m²/d) than in VF-CWs (−3.99 to −0.33 g/m²/d) (Figure 2), with a significant negative relationship to the DO and ORP value (Table 3). This indicates that anaerobic denitrification is proposed as the major route for removing NO₃-N in CWs (Pham et al. 2021).

The removal loadings of TSS and COD were ranked in the following order (Figure 2): high-HLR-CWs (15.34–19.24 g/m²/d of TSS and 36.47–99.89 g/m²/d of COD) > medium-HLR-CWs (8.55–11.02 g/m²/d of TSS and 18.22–53.16 g/m²/d of COD) > low-HLR-CWs (4.10–5.66 g/m²/d of TSS and 9.04–28.33 g/m²/d of COD). This may be because higher HLRs bring more organic and inorganic nutrients for microbial growth, thus promoting the biodegradation of pollutants. Moreover, we also found that WT raised with the increase of HLRs in winter, possibly due to the CWs was set up in a field, so the WT value in effluent was often lower than in influent (Table 2). Therefore, the lower HLR is, the less wastewater is treated in CWs, and the lower WT is, which can just explain why the removal loading of pollutants reduced following the decrease of HLRs.

In addition, zeolite owns a strong capacity of ion exchange for NH₄-N, which was often used on the removal of NH₄-N from wastewaters. For example, Du et al. (2020) found that CWs with zeolite substrate performed well on removing NH₄-N, with the average removal efficiency of 98.2%. This is accordance with our findings, in which the TN removal in zeolite-CWs was higher in other wetland systems (Figure 2).

Generally, the presence of plants is considered to be beneficial to the pollutant removal, because plants cannot only absorb and uptake contaminants, but also function as a carbon supplier for microbe metabolism (Stefanakis & Tsihrintzis 2012). However, neither the plant presence nor the plant type plays a significant role in the purification of black-odorous water (Table 4). For one thing, the effect of plant absorbtion/uptake could be ignored compared with the adsorption of substrates; for another thing, the black-odorous water contained plenty of organic and nutrient substances (Table 2), which supplied sufficient carbon and energy source for microbial growth rather than the root exudates released from plants.

CWs are demonstrated to be effective for purifying black-odorous water. In warm summer, since the activity of microorganisms are stronger, microbial degradation is considered as the most important way for black-odorous water purification in CWs. In cold winter, microbial effect becomes weak but filter adsorption ability is enhanced, so the joint effect of biodegradation and adsorption is proposed to be the main removal pathway of contaminants in CWs. Flow type and HLR are the most significant factors impacting the purification result of black-odorous in CWs. Among them, flow type can decide the ORP condition of the systems, which affect the microbial community distribution and the bacterial degradation performance, HLR can change the retention time of pollutants in the systems, and can also regulate the WT to affect the microbial activity.

Above all, VF-CWs with high HLRs are believed as the optimal choice for purifying black-odorous water. On the one hand, VF-CWs have good oxygen transport capacity, which is beneficial to the aerobic degradation of organic matters and the nitrification of NH₄-N. The unsaturated design of VF-CWs can also extend the contact time between pollutants and filters, thus improving the adsorption effect of TSS and PO₄-P. On the other hand, higher HLRs can increase the total removal loading of CWs per unit time, which make more efficient potential use of CWs on wastewater purification. HLRs can also adjust the WT of CWs, particularly in winter, and therefore enhance microbial effect. Furthermore, use of zeolite as the filter of CWs can further promote the removal of NH₄-N via ion exchange.

CWs are demonstrated as effective to purify black-odorous water in both summer and winter. In which, biodegradation is the most important process of pollutant removal in warm season, while microbial degradation and filter adsorption jointly contribute to the contaminant treatment in cold season. Flow type and HLR play a significant role on the purification of black-odorous water in CWs. Flow types may affect the dominant bacteria distribution by deciding ORP conditions, and HLRs can impact the pollutant removal effect by adjusting their retention time. The combination of VF-zeolite-high HLR is proposed to be the optimal design of CWs for purifying black-odorous water. Therefore, VF-CWs with higher DO and ORP values are beneficial to aerobic process of contaminants, zeolite filters may notably promote the absorption of nitrogen, and higher HLRs may enhance the pollutant removals by prolonging the contact time and providing higher WT for microbial effect.

This study was jointly funded by National Natural Science Foundation of China (41907293), Natural Science Foundation of Guangdong Province, China (2022A1515011319), Graduate Science and Technology Innovation Fund of Zhongkai University of Agriculture and Engineering, China (KJCX2022021), and National College Students' innovation and entrepreneurship training program (202111347023).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.