A number of innovative nitrogen removal technologies have been developed to address the treatment challenges caused by stringent regulations and increasing chemical and energy cost. A major contributing factor to these challenges is the liquid stream originating from the process of dewatering anaerobically digested solids. This liquid, also knows as centrate, reject water or sludge liquor, can cause an increase of up to 25% in ammonia loading. The recently discovered anaerobic ammonia oxidation (anammox) process is a major breakthrough for treatment of these streams as it has the potential to remove up to 85% of nitrogen load without external carbon source addition. The anammox process is combined with another process that oxidizes half of the ammonia to nitrite (nitritation) in a separate reactor such as in the SHARON process, or in the same reactor such as in the DEaMmONification (DEMON) process. Despite intensive laboratory research for the last 10 years to fully understand these processes, there is still a high level of skepticism surrounding the implementation of full-scale systems. The reason for this skepticism could be due to frequent failures observed in the lab scale systems as well as reported slow bacterial growth. We think that this technology might be used more effectively in the future if process kinetics, inhibition and toxicity can be better understood.

This work focused on the DEMON process with a goal to understand the kinetics and inhibition of the system as a whole and the anammox process in particular. A DEMON pilot study was undertaken at the Alexandria Sanitation Authority (ASA) and had several study participants, including ASA, the District of Columbia Water and Sewer Authority (DCWASA), CH2M Hill Inc., Envirosim Ltd, the University of Innsbruck and Virginia Tech. We investigated the growth rate of anammox bacteria within a quasi-optimal environment. Laboratory-scale experiments were conducted to assess anaerobic ammonia oxidation inhibition by nitrite as well as aerobic ammonia oxidation inhibition by compounds present in the DEMON reactor feed, such as a defoaming agent, a sludge conditioning polymer, and residual iron from phosphorus removal practices.

The study revealed that the DEMON process can be efficiently controlled to limit nitrite accumulation capable of causing process inhibition. The target ammonium loading rate of 0.5 kg/m3/d was reached, and no upset was noticed for a loading up to 0.80 kg/m3/d with an HRT of 1.7 days. The ammonia removal efficiency reached an average of 76% while total nitrogen removal efficiency had an average of 52%. Most of the process upsets were caused by aerobic ammonia oxidation failure rather than anammox inhibition. Failure in ammonia oxidation affected pH control, a variable which is at the center of the DEMON process control logic. The pilot study is summarized in Chapter 3 of this Dissertation.

The low anammox maximum specific growth rate (µmax,An) as well as nitrite inhibition are historically reported to be the major process challenges according to the literature, but the degree to which each contributes to process problems differs widely in the literature. In this study, we estimated µmax,An by using the high F:M protocol commonly used for nitrifying populations. We also studied the effect of both short term and sustained nitrite exposure on anammox activity. In this study, µmax,An was estimated to be 0.017 h-1. The study results also suggest that anammox bacteria can tolerate a spike of nitrite-N at concentrations as high as 400 mg/L as long as this concentration is not sustained. Sustained concentrations above 50 mg/L caused a gradual loss of activity over the long term.

Finally, the inhibition of aerobic ammonia oxidizing bacteria (AerAOB) observed in the DEMON reactor was investigated using laboratory experiments and is reported in Chapter 6. AerAOB inhibition was, in most cases, the main reason for process upset. Compounds that were suspected to be the cause of the inhibition were tested. The study noticed that a defoaming agent, polymer and ferrous iron had some inhibiting properties at the concentrations tested.