Towards a general theory within wastewater treatment: computational and experimental examples utilising fundamental laws to describe microbial community structure in wastewater treatment systems

Herzyk, Tymon Alexander (2025) Towards a general theory within wastewater treatment: computational and experimental examples utilising fundamental laws to describe microbial community structure in wastewater treatment systems. PhD thesis, University of Glasgow.

Abstract

Wastewater engineering needs novel, sustainable systems designed to cope with the pressures of climate change, urbanisation and increasing water scarcity. Modern molecular methods in microbiology are allowing us to interrogate the complex microbial communities underpinning biological wastewater treatment technologies, which affords the opportunity to design and control them. The application of theory is a fundamental tool to achieve this, however, current theories, whilst numerous, are partial and idiosyncratic. A concerted research effort is required to generate generic, widely accepted theory for wastewater treatment design. This thesis approaches the problem of defining fundamental rules on the structure and dynamics of wastewater microbial communities from two different perspectives.

The first asks the question if there was a generic theory on community assembly for a wastewater microbial community, is it possible to identify its parameters using the type of data that is routinely collected? To answer this an empirical approach is used where a model is assumed, in this case the simple neutral model. It is applied to generate a range of synthetic relative abundance time series with a range of different features, such as different sampling frequencies. The model is then calibrated using these data under various scenarios to assess parameter identifiability. The process of efficiently simulating, sampling and calibrating was achieved by developing an open-source user-friendly and adaptable computational framework. The results demonstrate fundamental difficulties in identifying neutral model parameters, even when using idealised time series. This is due to misrepresentation of the expected variance and correlated parameters. As such, there is a need to independently determine some of these parameters a priori. Reduced sampling periods and frequencies are also shown to impact model calibration, leading to systematic errors in the estimates of certain parameters. The research highlights the need for increasing information within relative abundance time series and demonstrates that this can be achieved by inducing a perturbation on a system. Thus, producing time series during periods of non-equilibrium is shown to be beneficial. Finally, it is demonstrated that for real data sets neutral model parameters must be considered as "effective" parameters as the estimates obtained will likely reflect a myriad of complex phenomena.

The second perspective is mechanistic. Growth kinetics, known to drive inter-species competition, are explored through energy and thermodynamics. This is achieved experimentally by culturing two species of methanotrophs, Methylomonas methanica S1 and Methylosinus trichosporium OB3b, within an isothermal calorimeter. Comparisons are made between the heat dissipated by the species and how they partition the available carbon between energy-yielding and biosynthetic reactions. The methodology developed for measuring heat dissipation is novel in its application to methanotrophic bacteria. These measurements identify that a significant amount of heat is dissipated during the growth of methanotrophic bacteria. Differences in growth are also observed with Methylosinus trichosporium producing more CO2 at the cost of a reduced biomass yield when compared to Methylomonas methanica. This difference in carbon partitioning is shown to be linearly related to the heat dissipated per unit of biomass, however, the impact this has on providing a competitive advantage is unclear. It is speculated that additional biochemical and physical phenomena need to be quantified in order to resolve the relative importance of thermodynamics on the different kinetics.

Simple models with parameters and variables that are easily understood and manipulated by engineers have a long history of being successfully deployed in engineering design. In this thesis, two approaches have been explored to describe the assembly and kinetics of microbial communities. It is shown that identifying model parameters for even the simplest descriptions of community assembly, neutral models, is fraught with conceptual and practical difficulties. Notwithstanding this, their ability to capture phenomena is apparent, and the challenge for engineering design is mapping the calibrated parameters onto characteristics of engineered systems. A more mechanistic approach attempts to use energy and thermodynamics to explain kinetics using methanotroph species as an example. Energy based methods are the basis of many models used in engineering design because they can capture physical phenomena while eschewing much of the underlying complexities in physical or chemical processes. The thesis falls short of delivering a generic rule on the relationship between energy dissipated and kinetics. However, the empirical data generated are already being used in the design of technology for heat recovery from wastewater treatment. In general, the thesis demonstrates the merits of pursuing simplicity in models for generic application in environmental biotechnologies.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: T Technology > T Technology (General); T Technology > TD Environmental technology. Sanitary engineering
• Colleges/Schools: College of Science and Engineering > School of Engineering
• Supervisor's Name: Sloan, Professor William and Gonzalez Cabaleiro, Dr. Rebeca
• Date of Award: 2025
• Depositing User: Theses Team
• Unique ID: glathesis:2025-85429
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 03 Sep 2025 10:43
• Last Modified: 03 Sep 2025 10:47
• Thesis DOI: 10.5525/gla.thesis.85429
• URI: https://theses.gla.ac.uk/id/eprint/85429