Simulation-based framework for design and optimization of wastewater treatment plants

Research output: Book/ReportPh.D. thesisResearch

54 Downloads (Pure)


Mathematical models have become increasingly vital tools in decision-making processes in engineering and beyond. Fueled by the continued growth in computing power and the recent boom in available data, these models have substantially grown in size and complexity, allowing for remarkably accurate simulations of highly complex process phenomena. Notwithstanding the progress, much of what is now achievable through these simulation models still remains not being fully utilized by industrial practitioners, especially during the early stage design of complex engineering systems. A good example is wastewater treatment plants, whose design problem presents a formidable challenge to engineers, and is often addressed through conventional design approaches that are based on industrial practices and experiences rather than mechanistic models. However, such approaches fall short of addressing the heightened level of complexity added to the design problem due to the recently shifting design objectives (e.g., increased focus on resources recovery), the rapidly growing number of emerging treatment technologies along with the uncertainties surrounding the influent compositions as well as the effluent quality requirements. Among the design professionals facing this challenge, there is, therefore, an increasingly pronounced need for integrating recent simulation and optimization techniques into the practiced design approaches.

The objective of this project was to address this need by investigating and developing systematic methodologies and tools to support simulation model-based design and optimization of wastewater treatment plants. The core question to be answered in this thesis is, how can we make the best use of the improved process understanding achieved in the form of high-fidelity process models in deciding a plant design and its operation in order to maximize its performance metrics? Through the use of fundamental process systems engineering tools and methods, the thesis systematically addresses this question by presenting a new comprehensive framework for simulation-based synthesis, design, uncertainty quantification, and optimization of wastewater treatment plants. This unified framework effectively integrates the first principle models of alternative wastewater processes with its newly developed methods and algorithms for formulating and solving plant synthesis, design space exploration, and optimization under uncertainty problems. Furthermore, the generic features of the proposed framework have been implemented in a set of new software tools to facilitate their wider adoption by non-specialist practitioners in other domains. The thesis also presents these framework tools (along with their application examples), which include SPDLab (for assisting with plant layout selection), easyGSA (for machine learning-assisted design space exploration based on global sensitivity analysis), and MCSKopt (for stochastically constrained simulation-based optimization under uncertainty using Monte Carlo simulation). Finally, the thesis highlights the capabilities of the proposed framework and its tools in several case studies derived from different design and uncertainty scenarios typically encountered in wastewater treatment plants.
Original languageEnglish
Place of PublicationKgs. Lyngby
PublisherTechnical University of Denmark
Number of pages202
Publication statusPublished - 2020


Sustainable Process Synthesis and Design

Al, R., Sin, G., Gernaey, K. V., Zubov, A., Alsina, X. F., Niedenführ, S. & Pantelides, C. C.

Marie Skłodowska-Curie actions


Project: PhD

Cite this