Remoção de poluentes em um sistema alagado construído de escoamento vertical com alimentação em pulso

Abstract

This study investigated the behavior of a vertical-flow constructed wetland system treating raw municipal sewage of approximately 100 population equivalents (around 1 m2/inhabitant). There were three units in parallel, two of which are investigated here. One of the units was planted with tifton 85 (Cynodon spp) and the other was unplanted (control unit). Each unit was fed with raw sewage for 2.5 days and had a rest period of 4.5 days. The influence of plants, feeding mode and operation time was investigated in terms of the removal of several physico-chemical parameters (pH, temperature, dissolved oxygen, BOD, COD, TSS, TKN, N-NH4+, nitrite, nitrate and total phosphorus). Also, the solids attached to the filter medium were collected at different depths (10 cm and 30 cm) for the evaluation of solids retention and most probable number of nitrifying and denitrifying bacteria. The effluent discharges were measured by volumetric method. Tifton 85 nutrient extraction capacity was evaluated. To this data were added the results from a previous monitoring, so the total time of the study comprised two and a half years. Two different hydraulic loading rates and dosing frequencies were tested for each filter in order to evaluate the impact of the perational mode on their performance. The hydraulic loading rate (HLR) applied in each bed from March 2010 until December 2011 was of 0.38 m3.m-2.d-1(batch of 940L of wastewater every 2 hours) Phase 1 and from January 2012 to July 2012 it was 0.45 m3.m-2.d-1(batch of 560L of wastewater every 1 hour) Phase 2. The planted unit performed better than the unplanted one, indicating a positive influence of the presence of plants, and its mean removal efficiencies for phases 1 and 2 were, respectively: BOD: 79 and 82%; COD: 72 and 81%; TSS: 83 and 85% ; TKN: 55 e 56%; Total-P: 45 and 21%. For the unplanted unit the results for each phase were: BOD: 78 and 80%; COD: 72 and 81%; TSS: 77 and 77% ; TKN: 46 and 47%; Total-P: 50 and 33%. Higher total and volatile solids attachment and higher amount of bacteria involved in the nitrogen cycle were observed in the planted filter medium, what can explain its higher nitrification and solids removal. The unplanted unit showed higher peak discharges. The application of smaller influent volume with a higher batch frequency improved the performance of both units, but a higher improvement was observed in the unplanted unit. No signs of medium clogging have been observed in the units. The system simplicity and the good removal efficiencies indicate its high applicability in small communities in Brazil.