Key points are not available for this paper at this time.
The photovoltaic industry has shown vigorous growth over the last decade and will continue on its trajectory to reach terawatt-level deployment by 2022–2023 and an estimated 4.5 TW by 2050. Presently, its elaboration is driven primarily by cost reduction. Growth will, however, be fuelled by the consumption of various resources, bringing with it unavoidable losses and environmental, economic, and societal impacts. Additionally, strong deployment growth will be trailed by waste growth, which needs to be managed, to support Sustainable Development and Circular Economy (CE). A rigorous approach to quantifying the resource efficiency, circularity and sustainability of complex PV life cycles, and exploring opportunities for partially sustaining industry growth through the recovery of high-quality secondary resources is needed. We create a high-detail digital twin of a Silicon PV life cycle using process simulation. The scalable, predictive simulation model accounts for the system's non-linearities by incorporating the physical and thermochemical principles that govern processes down to the unit operation level. Neural network-based surrogate functions are subsequently used to analyse the system's response to variations in end-of-life and kerf recycling in terms of primary resource and power consumption, PV power generation capacity, and CO2 emission. Applying the second law of thermodynamics, opportunities for improving the sustainability of unit operations, the larger processes they are the building blocks of, and the system as a whole are pinpointed, and the technical limits of circularity highlighted. We show the significant effects changes in technology can have on the conclusions drawn from such analyses.
Bartie et al. (Tue,) studied this question.