The ability of several mesocosm-scale and full-scale constructed wetlands (CWs) to remove pharmaceuticals and personal care products (PPCPs) from urban wastewater was assessed. The results of three previous works were considered as a whole to find common patterns in PPCP removal. The experiment took place outdoors under winter and summer conditions. The mesocosm-scale CWs differed in some design parameters, namely the presence of plants, the vegetal species chosen (Typha angustifolia versus Phragmites australis), the flow configuration (surface flow versus subsurface flow), the primary treatment (sedimentation tank versus HUSB), the feeding regime (batch flow versus continuous saturation) and the presence of gravel bed. The full-scale CWs consisted of a combination of various subsystems (ponds, surface flow CWs and subsurface flow CWs). The studied PPCPs were ketoprofen, naproxen, ibuprofen, diclofenac, salicylic acid, carbamazepine, caffeine, methyl dihydrojasmonate, galaxolide and tonalide. The performance of the evaluated treatment systems was compound dependent and varied as a function of the CW-configuration. In addition, PPCP removal efficiencies were lower during winter. The presence of plants favoured naproxen, ibuprofen, diclofenac, salicylic acid, caffeine, methyl dihydrojasmonate, galaxolide and tonalide removal. Significant positive correlations were observed between the removal of most PPCPs and temperature or redox potential. Accordingly, microbiological pathways appear to be the most likely degradation route for the target PPCPs in the CWs studied.
Keywords: constructed wetlands, pharmaceuticals