BACKGROUND: In this study, an onsite pilot sequencing batch reactor (SBR) of 450 L was operated at a local Flemish slaughterhouse in two consecutive periods. During period I, nitrogen removal via nitritation/denitritation (Nit/DNit) was achieved by the use of an innovative startup strategy without seed sludge and subsequently maintained by the use of aeration phase length control based on the oxygen uptake rate (OUR). In period II, the operational strategy was modified. In order to obtain Nit/DNit granular sludge (GS), an anaerobic (feast) feeding was introduced followed by a dynamic controlled (OUR-based) aerobic (famine) reaction step. RESULTS: After 74 days, the local (Flemish) legislation standards for COD (125 mg O 2 L −1 ) and T-N (15 mg N L −1 ) were fulfilled. During period II, floccular sludge was partially transformed into GS, which resulted in increased COD and NH 4 -N conversion rates of 50 and 60% respectively. Average removal efficiencies for SCOD, N and P during period II were 89.5 ± 4.2, 88.8 ± 9.8 and 84.4 ± 9.3% respectively. Despite the relatively low biomass concentrations (2.5 g L −1 ), excellent sludge settling was reported, with SVI 30 < 100 mL g −1 from day 167 onwards. Moreover, it was found that the operational strategy and temperature variability had a crucial impact on the microbial community dynamics and granule growth.CONCLUSION: This study demonstrated the Nit/DNit BNR-GS system in a real operational environment through the operation of an onsite pilot plant.

BACKGROUND This study investigated the use of a novel start‐up strategy to obtain nitrogen removal via nitrite using real‐time aeration control, in a sequencing batch reactor (SBR1), for the treatment of slaughterhouse wastewater. In addition, another SBR (SBR2) was operated with the same objective, but with a start‐up from seed sludge. Furthermore, analysis of the nitrous oxide (N2O) production and emission was conducted for different nitritation levels. RESULTS Through the absence of an inoculum in SBR1, the nitrifying population could be regulated from the beginning. Beceause of this, nitrite oxidising bacteria could be prevented. This was proved by the nitritation degree (ND) which was already above 80% from day 28, although it took 59 days to achieve sufficient nitrogen removal efficiency (> 80%). In contrast for SBR2, it took 90–110 days to achieve full nitritation, however the nitrogen removal efficiency was above 90% at all times. The maximum N2O emission resulted to be 0.12% of the nitrogen load. CONCLUSION This study suggested the importance of seeding sludge for the achievement of full nitritation. Moreover, a strong positive relationship between the ND and N2O emission was observed. © 2017 Society of Chemical Industry