Upgrading Urban Wastewater Management by Integrating Microbial Processes: Stoichiometric, Kinetic and AI-based Modelling

Abstract

Rapid urbanization is increasing wastewater generation in cities worldwide, placing growing pressure on urban sewer systems. Many sewer networks and wastewater treatment facilities, however, were constructed decades ago and have undergone limited modernization. As cities adopt increasingly ambitious sustainability goals, the combination of rising wastewater loads and ageing infrastructure poses significant challenges for sewer system management. Sulphide‑induced corrosion, for example, progressively degrades sewer pipes, leading to substantial maintenance costs. At the same time, higher wastewater volumes and more stringent effluent standards are increasing energy consumption and greenhouse‑gas emissions from treatment processes. Addressing these challenges requires a systematic upgrade of urban sewer management to reduce costs and improve environmental performance.

Although urban sewer systems are large, heterogeneous, and spatially distributed, their management can be broadly organised into three domains: source management, sewer networks, and wastewater treatment. At the source, cities lack robust, city‑specific tools to evaluate policies such as the integration of food waste into urban wastewater systems. Within sewer networks, hotspots of hydrogen sulphide generation, infrastructure corrosion, and associated health risks remain difficult to identify and manage. At wastewater treatment plants, effective and reliable process control continues to be challenging.

Despite their differing locations and scales, these challenges share a common foundation: the regulation of microbial activity throughout the sewer system. Wastewater management is therefore fundamentally a problem of managing microbial processes rather than isolated infrastructure components. In this work, we integrate microbial process knowledge across source management, sewer networks, and treatment processes to address these challenges and improve the performance and sustainability of urban sewer systems.

For source management, we develop a city‑scale Urban Biowaste Flux model based on a stoichiometry‑based life‑cycle assessment approach. The model quantifies material flows, economic costs, and greenhouse‑gas emissions under scenarios with and without food‑waste integration into sewer systems. It is calibrated using data from Hong Kong and applied to 28 cities worldwide. The results indicate that net system costs increase linearly with food‑waste moisture content and reveal a threshold of approximately 50 kg/capita/yr, above which food‑waste integrated system becomes cost‑effective. Optimized management strategies can reduce greenhouse‑gas emissions by up to 69% compared with conventional waste‑separation approach.

For sewer‑pipe management, we develop a high‑resolution sewer‑network model that links hydraulics, biofilm kinetics and concrete‑corrosion chemistry. Application to nine catchments in Hong Kong reveals that 54% of pipes exceed the 7 ppm hydrogen sulphide safety limit, 3.4% exceed 100 ppm, and 1.25% face high corrosion risk. This corresponds to about 676 km of pipes at risk of failure by 2030. Using the model to guide targeted oxygen dosing reduces overall hydrogen sulphide levels by 21% and cuts the length of high‑risk pipe segments by 79%.

For wastewater treatment, we develop a hybrid modelling framework that combines mechanistic Monod kinetics with a physics‑informed neural network to learn unknown microbial reaction kinetics from limited and noisy data. Tested using simulations, batch experiments and continuous‑flow systems, the model achieves R² values of 0.98–0.99. It consistently outperforms purely mechanistic and purely data‑driven models and enables optimal process control.

By integrating microbial process knowledge across source, sewer and treatment domains, this study provides a practical and transferable framework for reducing greenhouse‑gas emissions, corrosion risk and operating costs in urban sewer systems, and supports the development of more resilient and low‑carbon urban sanitation.

Date of Award	2025
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Guanghao CHEN (Supervisor)