Nitrogen Removal in Woodchip-based Biofilters of Variable Designs Treating Agricultural Drainage Discharges

Subsurface flow wetlands with a woodchip-based biofilter are known to act as effective measures in mitigating site-specific nutrient losses in agricultural drainage. To optimize operation, the combined effect of water temperature, hydraulic residence time (HRT), and filter design is fundamental. This 2-yr study evaluated a combination of three flow designs and two woodchips mixtures at a test facility receiving agricultural drainage discharge from a 78-ha catchment. The biofilter test facility consisted of six independent parallel-constructed wetlands (CW1-CW6), with two types of filter mixtures and three different hydraulic designs (horizontal, vertical upward, and vertical downward flow). Horizontal CWs performed best with an annual removal of total N (TN) amounting to 53 to 54% of the load, corresponding to a daily removal of 2.16 to 2.32 g N m(-3) d(-1). Removal by the vertical CWs amounted to 38 to 60% of the TN load or 1.19 to 2.31 g N m(-3) d(-1). Despite low temperatures and HRT down to 3 h, all CWs functioned well in winter with 20 to 30% N removal. In contrast, during summer when the hydraulic load was low and HRT was up to >96 h, N removal was similar to 100%, and the reduction sequence proceeded with sulfate reduction and formation of H2S. Statistical analysis and model development revealed that the percentage removal of both TN and nitrate N was mostly dependent on water temperature and HRT, explaining 81.7 to 91.1 and 83.5 to 89.5% of the variation, respectively.