Nutrient removal, methane and nitrous oxide emissions in a hybrid constructed wetland treating anaerobic digestate

A pilot hybrid constructed wetland (CW) planted with reeds (Phragmite australis) and rice (Oryza sativa L.) was designed to treat liquid anaerobic digestate in the Yangtze River Delta, China. The hybrid CW system was composed of four stages: two reed vegetated vertical subsurface flow beds (VSSF: U1 and U2) in sequence, followed by a reed vegetated horizontal subsurface flow bed (HSSF: U3) and a rice vegetated surface flow bed (SF: U4). The average loading rate of digestate was 3.6 m3 per day during the experimental period. The average concentrations of TN and TP in the influent were 379 ± 58 mg L−1 and 29.6 ± 9.2 mg L−1, while the average removal efficiencies of TN and TP were 94.6% and 88.4%. Both TN and TP removal efficiencies in the second VSSF containing zeolite gave the highest removal performance, in which the mass removal rates were 21.3 ± 8.0 g-N m−2 d−1 and 0.99 ± 0.69 g-P m−2 d−1, respectively. Similarly, the highest removal performance for COD was also observed in the second VSSF with a mass removal rate of 79.9 ± 72.4 g-COD m−2 d−1. On the other hand, the average CH4 and N2O fluxes were highest in the first VSSF, at 31.8 ± 12.9 mg m−2 h−1 and 3.7 ± 2.8 mg m−2 h−1, respectively. There was a significant linear relationship between CH4 flux and DOC concentration in the pore water as well as a correlation between N2O flux and TN concentration. Total GWP of the hybrid CWs, derived from CH4 and N2O emissions, was 792.4 kg CO2-eq, of which CH4 and N2O emissions accounted for 66.0% and 34.0%, respectively. Consequently, the hybrid CWs emitted on average 0.93 kg CO2-eq to remove 1 kg COD while the average EF of TNin was 0.34%, suggesting that the use of multistage hybrid CWs could be efficiency-wise and environmentally a promising strategy for anaerobic digestate treatment. © 2020 Elsevier B.V.