Sponge-bed trickling filters for nitrogen removal from anaerobically treated sewage : mechanistic insights and practical experiences

Subtitle:Sponsgebaseerde oxidatiebedden voor stikstofverwijdering uit anaeroob gezuiverd rioolwater : mechanistische inzichten en praktijkervaringen

The Latin American region still struggles with basic sanitation problems that developed nations have long surpassed. Hence, proper technologies should be considered for increasing sewage treatment coverage, entailing process robustness, operational simplicity, and low capital and operational expenditures. In this way, anaerobic sewage treatment has been widely applied in the region, mainly in Brazil, fulfilling a major role of organic carbon abatement. As an anaerobic process, intrinsic limitations thus exist concerning nitrogen removal. Among the consolidated post-treatment options, trickling filters have been extensively applied due to their remarkable effluent quality in terms of residual organic carbon removal. Trickling filters are typically filled with rocks as a support material for the attached biomass growth (hereafter termed rock-bed trickling filters). Research has been done on improving the performance of such reactors for carbon and nitrogen removal by replacing rocks with a sponge-based support media, the so-called sponge-bed trickling filters (SBTFs). A core advantage of the SBTFs lies in the high capacity of biomass retention, allowing for the colonization of slow-growing bacteria, such as those involved in the nitrogen cycle. Moreover, implementing SBTFs reduces land requirements and possibly saves construction costs compared with rock-bed trickling filters, as secondary settlers can potentially be eliminated. This doctoral research work focuses on the use of sponge-bed trickling filters for the post-treatment of anaerobic effluents. The overall goal is to establish highly efficient nitrogen removal, dealing with residual organic carbon and integrated with the abatement of dissolved gases in the anaerobic effluent. Demo-scale experimental studies are thus combined with mathematical modelling and simulation work. Chapter 1 gives a general introduction, outlining consolidated technologies for the post-treatment of UASB reactor effluents. Research challenges are summarized considering the (i) design of SBTFs following anaerobic sewage treatment, (ii) the relevance of influent characteristics, (iii) the need for long-term experimental assessments, and (iv) nitrogen removal pathways at the core of this thesis, namely: conventional nitrification-denitrification and partial nitritation-anammox (PN/A). Additional autotrophic nitrogen removal processes are outlined based on the link with dissolved gases abatement (i.e., denitrifying anaerobic methane oxidation (DAMO) and sulfur-based denitrification (SBDN)). In Chapter 2, a literature review is carried out on the design and operation of trickling filters post-UASB reactors. Practical experiences are critically compiled to derive the most important design criteria and relevant influent characteristics to predict process performance for organic matter removal. An outlook is given on process configurations for improving nitrogen removal via heterotrophic denitrification or partial nitritation-anammox. Based on the consolidated design criteria, a demo-scale SBTF was built and operated under the scope of Chapter 3, which deals with the relevance of influent characteristics, most specifically inorganic carbon limitation during nitrogen conversions. A 300-day monitoring campaign showed that a lack of inorganic carbon impaired nitrification, and nitrite oxidizing bacteria (NOB) were less affected. Bicarbonate was added as a state variable to properly describe inorganic carbon limitation, and sigmoidal kinetics were applied. The resulting model was able to capture the overall experimental behaviour. In Chapter 4, the developed model was used to mechanistically assess the effect of key reactor and kinetic parameters controlling nitrogen conversions in SBTFs in the long-term. A simulation study was performed considering the key reactor-specific parameters that influence the formation of nitrogen gas via heterotrophic denitrification or anammox process, identified via sensitivity analysis. The results support that the interplay between the oxygen transfer coefficient, external mass transfer resistance, biofilm thickness, and specific surface area of the sponge-based support media influences the optimum oxygen concentration to sustain ammonium oxidizing bacteria (AOB) activity without compromising anammox bacteria growth. Particular attention was paid to process start-up, which was identified as a primary bottleneck. Standalone biomass inoculation strategies for promoting a fast partial-nitritation anammox are ineffective if inhibitory oxygen levels remain at the biofilm-liquid interface. Effluent recirculation coupled to sewage by-pass led to a quick establishment of heterotrophic denitrification; however, it was limited at approximately 55% total nitrogen removal. Overall, high performance for both processes (i.e., heterotrophic and autotrophic) relies on particular reactor-specific parameters. In Chapter 5, a long-term experimental comparative study was performed in two SBTFs operating in parallel following a UASB reactor treating real sewage. Effluent recirculation to the top compartment of the SBTF showed a rapid increase in total nitrogen removal efficiency. Nevertheless, further supplying organic carbon via sewage by-pass was detrimental to AOB activity. Stepwise ventilation strategies decreased volumetric ammonium conversion rates; however, nitrate remained the main end-product throughout the monitoring period, meaning NOB repression was ineffective. The model developed under Chapter 3 was further used to gain process insight based on the observed experimental data. Results indicated that dissolved gases in the anaerobic effluent likely hampered AOB during restricted ventilation of the SBTF. Furthermore, under the observed temperature range, the long-term ingrowth of anammox bacteria tends to be constrained. In Chapter 6, the fate of dissolved methane and H2S in the anaerobic effluent during nitrogen conversions in SBTFs was assessed. The developed model (Chapter 3) was expanded to account for stripping and biological conversion processes of dissolved methane and H2S. Simulations showed that nearly all gases were stripped from the anaerobic effluent at the top compartment of the SBTF. If a (partially) closed SBTF is applied, stripping is therefore decreased, and practically all methane and H2S were oxidized by methane oxidizing bacteria (MOB) and sulfide oxidizing bacteria (SOB), respectively. Nevertheless, total nitrogen removal efficiencies were impaired due to the competition for oxygen, which was generally lost by AOB. Simulations also did not sustain the occurrence of DAMO or SBDN. Further experimental tests with a closed SBTF fed with desorbed anaerobic effluents showed that nitrogen conversions are potentially better handled if methane and H2S are removed upfront. In Chapter 7, the main findings are revised involving both mechanistic insights and practical experiences. Perspectives concerning practical implications and research needs are provided, and take-home messages conclude this thesis.

ISBN:9789463574815

Publication year:2022

Accessibility:Embargoed