Advancing the Application of Modeling and Statistical Methods in the Research and Development of Environmental Technologies
Advisor: Associate Professor Jeremy S. Guest
The growing complexity of sustainability challenges has called for systems thinking and science-based decision making for technology research and development (R&D). This need is particularly pressing for the field of water and sanitation as the current rate of progress is deemed insufficient for the global fulfillment of the 6th sustainable development goal – universal sanitation for all by 2030. This dissertation addresses a critical barrier to the success of early-stage sanitation and resource recovery technologies through optimal design and deployment: a lack of consideration for interplay among design decisions, technology parameters, and context-specific characteristics under uncertainty. This barrier stems from a lack of established procedures to systematically characterize the expansive landscape of designs and deployment contexts under uncertainty, leading to the inability to elucidate critical barriers, tradeoffs, and research opportunities to advance the sustainability of environmental infrastructure. Computational modeling, uncertainty analysis, and sensitivity analysis, as individually established tools, have been recognized as important elements for the simulation and result interpretation step of model-based system design or optimization. However, a structured approach to the use of modeling tools together with statistical methods for R&D prioritization has yet been synthesized.
To this end, this work: (1) through rigorous global sensitivity analyses of a sunlight-mediated waterborne pathogen inactivation process, elucidated the relative importance of modeling assumptions and the impact of interplay among design decisions, technology parameters, and context-specific characteristics on technology performance; (2) through uncertainty and sensitivity analyses of a multi-scale process model, identified the technological drivers and quantitatively delineated the R&D targets of encapsulated anaerobic technology for sustainable distributed treatment and energy recovery of high-strength industrial wastewater; and (3) developed and introduced an open-source Python package, QSDsan, to integrate and streamline process modeling, unit design, system simulation, and sustainability assessments of sanitation and resource recovery technologies under uncertainty. Altogether, this research demonstrates the synergistic application of modeling and statistical methods to expedite technology R&D and advances this approach by providing a structured, flexible, and transparent computational tool.