Advancing the Circular Economy of Critical Materials

Clean energy technologies, which aim to reduce environmental impacts from production and consumption, are gaining traction in an era of ever-increasing environmental awareness. As industries around the globe transition to more environmentally friendly approaches, a greater emphasis has been placed on critical materials (especially rare earth elements) that provide essential and specialized properties to help realize clean energy products or systems (e.g., electric vehicles and wind turbines). Due to increasing demand and substantial supply risks, the market volatility of critical materials (such as price fluctuations and spikes) has been aggravated in recent decades. To mitigate supply uncertainties for emerging clean energy technologies, a transition from the conventional linear economy (take-make-dispose) to a circular economy (CE) is expected to play a vital role. In this context, a circular economy considers end-of-life products and manufacturing waste streams as potential resource flows, from which material value may be extracted to create "new" products.

To support the basic science needs and provide decision-making insights for the nascent circular economy technologies being developed, growing attention has been drawn to crosscutting research. Crosscutting research develops theoretical, computational, and experimental methods that are necessary to gain a comprehensive understanding of the CE technologies under development, so as to identify improvement opportunities and facilitate their industrial implementation. In a market-driven society, techno-economic assessment (TEA) is gaining prominence in the domain of crosscutting research. The objective of TEA is to evaluate the technical viability and economic feasibility of a technology or system. Through production cost-benefit modeling and financial analysis, TEA can assist in making more economically informed decisions regarding process designs and facilitating the commercialization of more CE approaches.

In an effort to accelerate the clean energy transition, this dissertation explores the application of multiple analytical approaches to accelerate the development and deployment of novel technologies that support the circular economy of critical materials. Three knowledge gaps are identified through a literature review: 1) a lack of suitable methods to address unknowns/uncertainties in evaluating the economic feasibility of processes/technologies at early stages of development, especially due to insufficient experimental data; 2) insufficient comprehensive cost-benefit analysis frameworks based upon TEA results, for example, how to interpret TEA metrics and identify optimal operating conditions; 3) Inadequate modeling techniques for the planning of end-of-life infrastructures that are currently underdeveloped.

To close each knowledge gap above, this dissertation pursues the following three overarching goals: 1) Design forecasting and prediction models that can provide decision support in the early stages of project/system development; 2) Extend the utility and enhance the robustness of existing TEA frameworks; 3) Support process planning and system design for end-of-life infrastructures. To accomplish these goals, four technical tasks were undertaken. First, a dynamic price model based on market supply and demand is developed to predict price fluctuations of critical materials that may influence the validity of a TEA. This model makes an analogy between the price dynamics and the moving pattern of a mechanical oscillator. Second, to evaluate the economic feasibility of value recovery activities (e.g., recycling and remanufacturing) that are still being conceptualized, a machine learning model using proxy variables as inputs is developed to provide early-stage estimations. Third, a pioneering remanufacturing system for electric vehicles is proposed and modeled as a stochastic activity network, whose performance is evaluated and optimized in terms of both operation and economy. This network model can be applied in conjunction with simulation, optimization, and topology analysis, which can be adopted to support decision-making across a diverse range of production systems, especially with applications to remanufacturing. Finally, to provide more insightful decision-making guidelines in terms of improvement opportunities, existing TEA models are enhanced by incorporating statistical learning approaches. The invention of graphics-based software tools that promote the adoption and use of TEAs and streamline the procedures for economic evaluations is also discussed.

The main intellectual contribution of this dissertation is a multitude of innovative analytical frameworks and paradigms that can facilitate the implementation and deployment of circular economy approaches for critical materials. By utilizing models and techniques from a diverse range of disciplines (e.g., system dynamics, machine learning, network analysis, and statistical learning), this dissertation broadens the methodologies used in early-stage decision-making for novel CE technologies/processes in development and advances the state of knowledge in system planning and economic analysis.