Wastewater resource recovery via the Enhanced Biological Phosphorus Removal and Recovery (EBP2R) process coupled with green microalgae cultivation

Conventionally, the objective of wastewater treatment has been the elimination of organic and inorganic pollutants, such as nitrogen and phosphorus, from wastewater. Current research promotes a paradigm shift, whereby wastewater is considered not only as a source of pollution but also as a source of nutrients, fresh water and renewable energy.

This new approach redefines the conventional wastewater treatment plant (WWTP) as a biorefinery from where different streams are split, each of them rich in different resources. Since many wastewater treatment infrastructures were built about 30 years ago, there is an opportunity of including these novel technologies as part of the future retrofitting and enlargements of the plants.

Nevertheless, most of the proposed resource recovery strategies suffer from intensive use of chemicals or energy. In extreme cases, the environmental impact of the technology by itself completely counters the benefit of resource recovery. As an alternative, this thesis proposes a new fully biochemical resource recovery process, referred to as TRENS.

The TRENS consists of an enhanced biological phosphorus removal and recovery (EBP2R) process combined with a photobioreactor (PBR). The EBP2R process is operated at relatively low solid retention time (SRT). Hence the bulk of nitrogen is preserved as ammonium, which is the preferred nitrogen source for green micro-algal growth.

The effluent criterion for the EBP2R is set to meet the micro-algal nutrient requirements in terms of nitrogen and phosphorus. To this end a phosphorus-rich stream (referred to as P-stream) is diverted from the anaerobic phase of the EBP2R and combined with a nitrogen-rich stream (referred to as N-stream).

As a function of the SRT and the P-stream diversion rate, different nitrogen-to-phosphorus ratios (N-to-P ratio) can be produced, thereby meeting the nutrient requirements of different micro-algal species. Organic carbon oxidation is minimized due to the low SRT. Therefore, most of the organic carbon is incorporated to the sludge via microbial assimilation or storage and conveyed to the anaerobic digester for biogas production.

The fraction of nitrogen which cannot be recovered is removed via completely autotrophic nitrogen removal (CANR). First, a feasibility assessment of the EBP2R process as an algal culture media generator was carried out using continuous-flow and sequencing batch reactor (SBR) configurations. Systems were modelled using the activated sludge model 2d (ASM-2d).

Regardless of the process configuration, factors that can potentially limit nutrient recovery comprise the system SRT and the nitrate recirculated to the anaerobic phase/reactor. Additionally, continuous-flow EBP2R systems can suffer from phosphorus starvation in the aerobic reactors as a result of excessive P-stream diversion.

Furthermore, in continuous-flow mode, the P-stream diversion increases the aerobic SRT, while the system SRT is kept. Consequently, nitrifying bacteria can proliferate in the continuous system oxidizing ammonia to nitrate. Therefore, at high P-stream flow diversions polyphosphate accumulating organisms (PAOs) may be outcompeted by denitrifying bacteria.

The sequencing EBP2R yielded to higher phosphorus recovery than the continuous flow system. For each of the EBP2R configurations a control structure has been developed and tested using a set of dynamic influent disturbance scenarios. The sequencing EBP2R system was found to be sensitive to large input disturbances.

Special care should be taken when tuning the controllers for the sequencing EBP2R to avoid too aggressive control actions that can potentially destabilize the system. Under dynamic conditions, the sequencing EBP2R show better performance in terms of phosphorus recovery and effluent quality (i.e. optimal N-to-P ratio fed to the PBR) than the continuous flow system.

Second, two short SRT EBPR systems were implemented as laboratory-scale continuous-flow and SBR reactor systems. Both systems suffered from extreme filamentous bulking (sludge volume index, SVI>1000 ml/g). Via 16rRNA amplicon sequencing we identified Thiothrix as the main filamentous bacteria driving activated sludge settleability.

Thiothrix proliferated in the reactors when sulphate was reduced to sulphur reduced compounds, such as sulphide, by sulphate reducing bacteria (SRBs). Phosphorus removal was poor during the filamentous bulking event, which was a consequence of the interactions between SRBs and PAOs in the anaerobic phase.

SRBs can compete with PAOs for volatile fatty acids under anaerobic conditions. Additionally, sulphide can inhibit phosphorus release by PAOs. As a result, PAOs were washed out from the systems. Filamentous bulking was mitigated and phosphorus removal was restored by reducing the anaerobic SRT of the SBR.

However, this strategy failed when applied to the continuous flow system, where only the SVI could be improved. When extending the aforementioned studies to include the PBR, we identified the lack of a model suitable to describe resource recovery from wastewater via green micro-algal cultivation. Furthermore, neither of models published in literature were compatible to interface with ASM-2d.

Therefore, the third part of the PhD project focusses on the development of a process model for micro-algal growth and substrate storage kinetics (referred to as ASM-A). To facilitate the integration in already well-stablished simulation platforms for wastewater treatment, e.g., the Benchmark Simulation Models 1 and 2, ASM-A was implemented as an extension to the ASM-2d.

A set of experiments at different laboratory-scales (microbatch, 1-litre and 24-litre SBR) was designed to generate data for model identification. Furthermore, an independent data set was used for model evaluation. The ASM-A can effectively predict the algal biomass growth, as well as the ammonium and phosphorus concentrations in the bulk liquid and the microbial stored phosphorus.

Conversely, our results suggest that the maximum uptake rate parameter for nitrate can be significantly affected by culture history. Therefore the prediction of bulk nitrate concentration and the microbial stored nitrogen requires case-specific model calibration. Finally, the models developed in PhD project were used to provide data for the inventory of a life cycle assessment (LCA) of the TRENS system implemented in the Copenhagen area.

The LCA highlighted the benefits of recovering nutrients but also suggested that heavy metals can potentially impose a bottleneck when reusing water and nutrients from the used water. Overall, this thesis describes the early stage design of the TRENS system, where model-based studies and laboratory-scale experiments have been used to define the optimal process operation and address future research needs.