[Effects of Three Bioretention Configurations on Dissolved Nitrogen Removal from Urban Stormwater]

Multiple chemical forms of nitrogen in urban storm water make its management challenging. Three types of bioretention systems were constructed in 2015 with loamy sand as filter media, including a conventional freely drained bioretention (CB), a modified bioretetion incorporated a submerged zone (MB1), and a modified bioretention incorporating a submerged zone with woodchips addition (MB2). This study investigated the role of vegetation, the use of submerged zone and carbon addition in achieving co-optimized dissolved nitrogen removal in bioretention systems. Twelve bioretention columns were monitored over a 12-month period of dosing with synthetic storm water under varying hydrology and nitrogen loading rates. All the studied bioretention systems could achieve very good ammonia removal (more than 95%) at an average inflow ammonia concentration of (5.45±2.21) mg·L^-1. The filter media sorption, nitrification and plants uptake were the main removal pathways for incoming ammonia. The effluent nitrate concentrations of the CB, MB1 and MB2 were (4.04±2.64)mg·L^-1 (31.3%), (0.84±1.18) mg·L^-1 (85.7%), and (0.26±0.48) mg·L^-1 (95.6%), respectively, at the average inflow nitrate concentration of (5.88±2.32) mg·L^-1. The use of the native species P. alopecuroides, a submerged zone and woodchips addition could effectively decrease the effluent nitrate concentration, reduce the washout and achieve high nitrate removal. Both plants uptake and denitrification were the two major pathways for removal of inflow nitrate. Inflow magnitude, antecedent dry days and inflow nitrate concentration were the main factors influencing the effluent nitrate concentrations for the three bioretention systems. The results highlighted that the bioretention design of the native species P. alopecuroides incorporated a submerged zone with 10% woodchips addition could consistently and effectively remove storm water nitrate under hydrological regime and nitrogen loading rates.