Concomitant phosphorus and nitrogen removal from agro-industrial wastewater through Partial Nitritation/Anammox process

Sustainable phosphorus (P) management is crucial for safeguarding global food security and preventing soil and water bodies' quality deterioration. Recently, the Anammox process was proven to be feasible for P recovery as hydroxyapatite from urban wastewater (1). This work firstly assessed the feasibility of P recovery by treating supernatant from agro-waste anaerobic digestion. Such wastewater was characterized by ammonium nitrogen (N) and P content of 743±17 mgNH4+-N/L and 29±1 mgPO42--P/L, respectively, and a Ca/P molar ratio of 1.64±0.09. The Partial Nitritation/Anammox (PN/A) process was operated according to a single-stage configuration in a 2L sequencing batch reactor (SBR). The bioreactor was inoculated by ANAMMOX© granulated biocatalyst, and run with 6-hour working cycles (i.e., 4 cycles per day). The process temperature was controlled at 35ºC, and the dissolved oxygen concentration was maintained below 0.2 mgO2/L to preserve anammox from oxygen inhibition. The Nitrogen Loading Rate (NLR) was increased throughout the experimental period from 0.22 to 0.33 kgN/m3?d by lengthening the feeding phase, taking into account the nitrite and ammonium accumulation. Moreover, a set of experiments was performed to pave the way for a novel non-damaging separation protocol to recover active biomass and mineral P from the sludge. Granules sampled from the bioreactor were exposed to 37 kHz sonication frequency for different periods (i.e., 60, 30, 15 min) and the effect of ultrasonication on biomass was investigated by measuring the Specific Anammox Activity (SAA) through the manometric method (2). As a result of the increasing NLR strategy, biomass showed significant growth and SAA improvement during the 6 months of the bioreactor operation. Mixed Liquor Volatile Suspended Solids (MLVSS) increased from 4.80 up to 6.77 g/L, while SAA's final value was 0.39 gN2/gMLVSS?d. In the absence of an external calcium source, simultaneous nutrient removal consisting of ammonium conversion rate up to 0.28 kgNH4+-N/m3?d and P removal efficiency of ca. 34% were achieved. As P concentration started lowering in the bulk liquid, the MLVSS to Mixed Liquor Suspended Solids (MLSS) ratio decreased over time from xx% reaching a final value of 75%. This reduction is suggesting probable precipitation of inorganic compounds inside the granules. The resulting Ca/P ratio in the liquid phase varied from 1.47 to 1.61. Thus, Amorphous Calcium Phosphate (ACP) and/or Tricalcium Phosphate (TCP) have been identified as possible deriving precipitates since their constituent molar ratio (1.5) falls within the Ca/P range. Regarding the ultrasonication tests, the SAA resulted being negatively affected when extending the treatment duration. Nevertheless, after 15 min of granules sonication, the SAA showed a limited reduction, up to 15%, compared with untreated granules collected from the bioreactor. In conclusion, the PN/A was confirmed to be a promising process to achieve a sustainable and simultaneous removal of ammonium nitrogen and phosphorous from digested agro-industrial wastewater. Higher P removal may be achieved by increasing the Ca/P ratio with an external calcium source. Moreover, this work represents a first attempt for unveiling an efficient and non-damaging biomass-to-mineral separation. Re-inoculating microorganisms to the bioreactor would be a key driver for evaluating the effective P recovery potential, and eventually for considering the future scaling up of the technique.