Heterotrophic Flocular Biomass Aggregation on Partial Nitrification-Annamox Granular Biomass Reactor: Operational Strategy for Continuous Processes with High Removal Efficiencies

Partial Nitrification - Anaerobic ammonia oxidation (PN-A) process is the most effective biological treatment for nitrogen removal in systems with low carbon/nitrogen ratio. Commonly, the PN-A processes are operated with granular biomass, since the anaerobic ammonium oxidizing bacteria (anAOB) has a low duplication rate and aerobic and anaerobic metabolic pathways are needed for the conversion of ammonia to nitrogen gas. In addition, granular biomass allows to separate the solids retention time (SRT) from the hydraulic retention time (HRT). Then, the PN-A process can be operated with short HRT, less than one day. In addition, the heterotrophic biomass will develop in floccular form when the HRT of the system is larger than the heterotrophic specific growth rate. At present, there is no evaluation of the effect of the progressive decrease of the HRT in a continuous regime PN-A granular process on the heterotrophic floccular biomass and on the nitrogen removal.

Thus, this work was aimed to determine the effect of increasing the nitrogen loading rate (NLR) through the HRT in a continuous PN-A granular reactor with gas agitation.

Experimentation of the PN-A process was carried out in a granular biomass bubble column reactor (BCR) with 3 L of working volume. During the startup the reactor was fed with the effluent from the anaerobic digestion of poultry manure. The ratio of Chemical Oxygen Demand (COD)/ Total Ammonia Nitrogen (TAN) was always less than 0.7 in order to prevent an effect of an organic matter overload. The NLR was modified decreasing the HRT from 1.90 to 0.15 days, and keeping the influent concentration constant at 400 mg TAN/L. The effect of the HRT variation was evaluated through the reactor performance determination, nitrogen removal percentages and biomass characterization such as relative abundance of populations, specific activities, sedimentation rate, sludge volumetric index and size of the granules.

The evaluation of the NLR increase (and simultaneous reduction of HRT) showed effects in the biomass structure for the operation at HRT=0.15 d with a NLR=2.6 kgN/L∙d. The lowest possible HRT with a good performance was 0.40 d, with a NLR of 1.06 kgN/L∙d and organic loading rate (VCO) of 0.44 kgCOD/L∙d, with efficiencies of inorganic nitrogen, TAN and COD removal of 83%, 85% and 49%, respectively. After this condition, the reactor HRT was changed to 0.15 d and the system showed flotation of granules and biomass washout, lowering the efficiency of nitrogen removal until 20%. The biomass flotation was because of the development of heterotrophic biomass in the surface of the granules, generating a greater area and lower density. When increasing the HRT again until 1.5 d, the inhibition was reversible: the biomass recovery showed removal efficiencies for inorganic nitrogen, TAN and COD of 95%, 98% and 51%, respectively. Thus, an efficiency greater than the stoichiometric value was achieved probably due to the adhesion of denitrifying biomass to the granule that is capable of reducing nitrate to nitrogen gas using organic matter as an electron donor. This was verified with the fluorescence in situ hybridization (FISH) assays; during the time period from the startup to the end of the experiments, an increase of 9 to 19% of unidentified bacteria was observed. Unidentified bacteria correspond to bacteria, which are not aerobic ammonium oxidizing bacteria (aeAOB), anaerobic ammonium oxidizing bacteria (anAOB) or nitrite-oxidizing bacteria (NOB); thus, the increase probably is due to the presence of heterotrophic bacteria.

In this work, we defined values for the optimization of the PN-A process; very high and efficient nitrogen removal can be achieved, moving through different stages of HRT variation, until the biomass inhibition and its subsequent recovery is demonstrated, evidencing also the high acclimation of the PN-A biomass.

Acknowledgements: This study was made possible by FONDECYT (Chile) [Grant Number 1140491] and INNOVA (Chile) [Grant Number 15VEIID-45613].