Mapping the Structure and Function of Microbial Populations in Enhanced Biological Phosphorus Removal
: (Alternative Format Thesis)

  • Minh Nguyen Quang

Student thesis: Doctoral ThesisPhD


Enhanced biological phosphorus removal (EBPR) is an environmentally sustainable technology to achieve P-removal from wastewater using phosphorus accumulating organisms (PAO). The deterioration of EBPR is often attributed to the competition between PAO and glycogen accumulating organisms (GAO). Metabolic models are a promising tool in the search for conditions favouring PAO. However, they have been limited to validation and interpretation of past results, owing to the uncertainty in operational conditions, environmental conditions and wastewater characteristics with respect to the microbial community composition (MMC), kinetics and stoichiometry of the biochemical conversions, and between MMC and EBPR performance respectively. This thesis aims to address these uncertainties to promote metabolic models for use in process optimisation.

A framework for automated article selection was evaluated against the past 40 years of EBPR research and used survey input factors observed in different EBPR systems, including the kinetic and stoichiometric parameters, as well as the resultant PAO fractions and P-removal efficiencies. Uncertainty and global sensitivity analyses were used to determine the impact of uncertainty in parameters required to initialise metabolic models for PAO and GAO on the prediction of soluble substrate, biomass and intra-cellular reserve conversions. Lastly, the steady-state conversions were predicted via multiparametric perturbation to determine the relationship between the input factors, PAO fraction and P-removal efficiency.

Uncertainty in input parameters led to significant variance in model predictions. They may not be assumed fixed. Nevertheless, intra-cellular reserve, biomass and VFA concentrations were more influential. Thus, detailed characterisation of the intra-cellular PHA, glycogen and poly-P, as well as influent wastewater would result in the greatest reduction of model prediction variance.

Contrary to consensus, near complete P-removal was equally likely with PAO fractions between 30-90%, and was possible with PAO fractions as low as 10%. High P-removal efficiencies without PAO dominating the MMC were associated with pH above 7.2 and SRT below 16 days. Conversely, poor P-removal despite PAO dominance was associated with SRT, HRT and influent carbon to phosphorus ratios exceeding 21 days, 0.7 days and 40 C-mmol P-mmol-1l-1 respectively. EBPR performance may be managed via input factors more readily manipulated in full-scale plants than those highlighted in lab-scale.

We highlight future directions for metabolic model development to improve prediction accuracy, challenge the pre-requisite assumption of PAO dominance and indicate the conditions favourable to EBPR performance in full-scale systems.
Date of Award23 Mar 2022
Original languageEnglish
Awarding Institution
  • University of Bath
SponsorsEngineering and Physical Sciences Research Council
SupervisorAna Lanham (Supervisor), Tim Rogers (Supervisor) & Jan Hofman (Supervisor)


  • EBPR
  • PAO
  • GAO

Cite this