Photovoltaic circular economy driven by tradable recycling credits: Policy synergies in an agent-based model

The rapid deployment of photovoltaic (PV) systems worldwide will generate massive volumes of end-of-life modules, yet the economic viability of recycling remains insufficient without effective policy intervention. Existing analyses predominantly rely on static models with exogenous prices and representative agents, limiting their capacity to evaluate multi-instrument policy interactions. This study develops an agent-based model (ABM) in which manufacturers, PV owners, and recyclers interact through double auction mechanisms across three coupled markets (product, recycling credit, and waste), with endogenous price discovery and evolutionary dynamics. Eight policy scenarios combining environmental tax, extended producer responsibility (EPR), and technology subsidy are simulated across Monte Carlo runs and evaluated through ANOVA and Sobol sensitivity analyses. The two dominant instruments operate through nearly orthogonal channels: the tax raises technology adoption by 27.4% and manufacturer profit by 20.5%, while EPR increases recycled volume by 436.5% and cuts landfill by 42.4%. This separation means their combination yields a Pareto improvement for both manufacturers and recyclers simultaneously. When all three policies are implemented jointly, the scenario achieves the highest composite performance (0.772), with a complementary 12.5% output gain and 80.8% recycler profit increase beyond additive expectations. These results offer actionable guidance for designing synergistic recycling policies, and Sobol analysis confirms that the identified mechanisms originate from market structure rather than parameterisation choices.