The threat of climate change has accelerated the shift in production and consumption patterns towards more sustainable systems. The proposition of a circular bioeconomy has favoured the development of biodegradable plastics in fast-moving consumer goods, especially in food packaging applications. Since biodegradable plastics can be disposed and treated alongside organic waste, including municipal and commercial food waste, they can contribute towards closed-loop material and product flows. Consumers play a key role as enablers of these flows, provided they accept, understand, use and dispose of biodegradable plastics appropriately. Mapping and facilitating the role of consumer behaviour in tackling post-consumer biodegradable plastic waste is therefore pivotal in designing sustainable systems for these materials and for achieving a circular plastic economy.

To date, literature on pro-environmental behaviour, including recycling, reuse and the adoption of biodegradable plastics has focused on understanding these actions as individual circular behaviours. Adopting a systems approach, we first mapped the consumer behaviour chain that unfolds across stages of acquisition, use and disposal of biodegradable plastics. Informed by a systematic literature review and focus groups, we then developed a stratified framework to identify and structure factors (system elements) that influencing how consumers interact with biodegradable plastics, ultimately enabling or hindering the progression of the behaviour chain. Overall, 18 system elements (further split in 35 system sub-elements) were identified across 6 broad structural categories consisting of resources, intrinsic drivers, data, value, infrastructure and policy. The emerging model acted as a foundation for the design of a survey, as a means of moving from qualitative to quantitative insights. We aimed to explore the dynamic interactions that occur within the system and investigate how system elements interact with each other and how they influence consumer behaviour, focusing on behavioural intentions for circular disposal (which, in the context of this study, was defined as the intention to of dispose biodegradable food packaging waste in food/organic waste bins) and disposal behaviour more broadly.

The survey was conducted at two academic institutions, Imperial College London (United Kingdom) and the University in California, Davis (United States), and formed the basis of a comparative case study and network analysis, in which the role of contextual setting of each case study site on behavioural intentions was also investigated. Results suggested that the presence and consistency of the relevant waste management infrastructure, as well as knowledge of bioplastic terminology and of appropriate disposal routes, play a key role in enhancing circular disposal intentions (and, hypothetically, ultimate circular disposal). The University of California, Davis's unique contextual setting, characterised by comprehensive state and institutional policies on organic waste management, contributed towards a more developed and consistent organic waste management infrastructure (including for food and food packaging waste), which in turn led to a higher proportion of surveyed participants to state having access to a separate food waste collection scheme and to display circular disposal behavioural intentions.

Our approach enabled us to identify leverage points most effectively across the consumption phase of biodegradable plastics to inform design strategies aimed at supporting consumers as they transition towards more pro-environmental behaviour in the context of plastics sustainability and integrated human-physical systems.

Material flow analyses (MFAs) provide insight into supply chain level opportunities for resource efficiency. MFAs can be represented as directed graphs of nodes and connecting flows. Network structure uncertainty (i.e., the existence or absence of nodes or flows between nodes) is pervasive in MFA and can undermine the reliability of the reconciled flow predictions; however, network structure uncertainty is typically not acknowledged in MFA studies and has yet to be rigorously studied. This talk will investigate network structure uncertainty by proposing multiple candidate node-and-flow structures and using Bayesian model selection and averaging to identify the most suitable flow structures as well as quantifying the parametric uncertainty that incorporates model-form uncertainty. The committee of network structure candidates is formed using both expert advice (exploitation) and randomized network structure generation (exploration). The principles of optimal experimental design (OED) are also developed to plan data acquisition that reveals the best MFA network structure while minimizing data collection costs. Through this endeavor, we conceive new computational algorithms: an efficient categorical stochastic optimization procedure that incorporates our domain knowledge about the MFA network structures, and an information-theoretic OED criterion for model-selection entailing a novel triple-nested Monte Carlo estimator.

The above methods are demonstrated using a series of toy MFA models and then a case study on the 2012 U.S. steel flow. A total of 16 candidate steel flow structures are proposed based on the connectivity of 4 different flows. With Bayes factor applied to estimate the probability of each candidate structure, parametric uncertainty using Bayesian averaging on a certain flow is estimated. Similarly, OED is applied to identify the best source of data to reduce the network structure uncertainty based on the 16 candidate structures. By applying this method, the network structure uncertainty of MFA can be addressed to further increase the value of MFAs for future policy making.

Communities and businesses continue to develop plastic recycling infrastructure including for the collection, sorting, purifying, and conversion into new recycled-content products. To promote sustainable infrastructure and systems a previous study analyzed polyethylene terephthalate (PET, Plastic #1) [1], while this effort examines polypropylene (PP, Plastic #5). PP represents a greater challenge to study than PET because data is relatively sparse. A material flow analysis of PP is used to describe product flows and the amounts and compositions of recycled bales of plastics. In addition to flows of products and the fractions recycled, mismanaged PP waste is approximated. Further processes for mechanical and chemical recycling of PP are modeled and substantially complete resource use and releases to the environment are determined. Results describe current PP recycling in the U.S. which can be used in scenario analyses, life cycle assessments, and in a presented discussion of opportunities to improve the systems involved.

[1] Smith, R.L., Takkellapati, S., and Riegerix, R.C. (2022), “Recycling of Plastics in the United States: Plastics Material Flows and Polyethylene Terephthalate Recycling Processes,” ACS Sustain. Chem. Eng., 10.1021/acssuschemeng.1c06845

The views expressed in this presentation are those of the authors and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency.

A sustainable source of rare earth elements (REEs) is essential for further implementation of many clean energy technologies (e.g., Nd and Dy for magnets in wind turbines). Conventional REE production extracts REE-rich minerals from the earth and separates them with high environmental impact. A new sustainable source of REEs that utilizes a highly selective and less environmentally impactful separation is needed. Phosphogypsum (PG), a toxic metal filled wastewater slurry from the production of fertilizers, is abundantly available and contains 0.05-0.5 wt% REEs. This presents an opportunity for the recovery of REEs and the remediation of the PG wastewater. A preliminary life cycle assessment (LCA) study has shown that PG remediation methods are effective at reducing the environmental impact of PG. Most prior work on REE recovery from PG aims to increase the leaching efficiency of REEs from the solid PG matrix. Little work has been done to assess the overall environmental impact and economic feasibility of a process using these leaching technologies. Here we use an integrated LCA, techno-economic analysis (TEA), and uncertainty/sensitivity analysis in Python to examine the effect of technological leaching parameters (temperature, acid concentration, and leaching time) on system-level environmental and economic feasibility. Indicators of global warming potential (GWP) and net present value (NPV) were chosen. A process for REE recovery from PG was developed based on literature PG leaching data and conventional REE production pathways primarily consisting of sulfuric acid leaching and chemical precipitation. A ‘black box’ separation was used to model separation of REEs into their individual elements at high purity. Preliminary analysis suggests that this system is profitable and has a comparable GWP to conventional REE production and PG remediation. Global sensitivity analysis shows that GWP is most sensitive to leaching temperature and NPV is most sensitive to leaching time and acid concentration. Additionally, through global sensitivity analysis, the optimum operating conditions for the leaching unit can be identified. Future work will examine the dependence of these results on different production capacities and the REE content in PG. Additionally, uncertainty due to the modelling of the black box separation will be quantified through wet lab experiments and uncertainty analysis. In summary, an open-source, integrated system analysis approach under a Monte-Carlo framework, can link process design parameters with sustainability metrics, enabling pre-optimization of the process and evaluation of the sensitivity and uncertainty of design parameters on process sustainability indicators. The results of this analysis can be used to prioritize research targets and understand trade-offs across the entire value chain (from waste feedstock to REE). Future work examining the entire landscape of possible process design decisions will provide much needed context for future REE recovery from secondary (waste) sources.