This preprint (arXiv:2603.25751), not yet peer-reviewed, presents a multi-fidelity modeling framework for the tritium fuel cycle using the open-source PathSim/PathView platform. The authors combine three approaches within one dynamic simulation: a zero-dimensional residence-time model to replicate an ARC-class fusion power plant's overall behavior, an intermediate-fidelity one-dimensional ODE model of tritium extraction in a liquid-metal bubble column reactor validated against existing literature, and high-fidelity multi-dimensional finite-element simulations using the FESTIM code that capture complex transport across material interfaces.

As a purely modeling study, it includes no experimental sample size; validation relies on published data rather than new measurements. Key limitations acknowledged include the framework's demonstration nature and the need for future coupling with neutronics and fluid dynamics tools.

Tritium self-sufficiency remains one of the most stubborn barriers to commercial fusion. Deuterium-tritium fusion consumes tritium rapidly, yet global supply is limited and expensive. Reactors must breed replacement tritium from lithium in the blanket via neutron capture while minimizing losses from permeation, inventory buildup, and inefficient extraction. Traditional system-level models often use oversimplified residence times that miss real physics; high-fidelity models are too slow for full-plant optimization. This work bridges that gap.

The paper goes further than most coverage by demonstrating consistent coupling across fidelity levels in a single environment, but it stops short of applying the framework to full design optimization or uncertainty analysis for specific machines. What earlier reporting on ARC-class reactors (Sorbom et al., Fusion Engineering and Design, 2015) missed was the detailed tritium extraction and transport physics that can make or break whether the tritium breeding ratio stays above 1.0 under realistic operating conditions. Similarly, earlier FESTIM-focused studies examined local transport but rarely fed results back into whole-plant fuel-cycle dynamics.

This physics-informed workflow therefore represents a genuine step toward digital twins for fusion fuel cycles. By revealing how design choices in the blanket, coolant, and tritium extraction systems affect overall inventories and breeding, it directly addresses a critical engineering barrier that plasma physicists alone cannot solve. If extended with uncertainty quantification and real experimental calibration, such tools could accelerate the timeline for tritium-self-sufficient pilot plants by allowing designers to identify and fix bottlenecks before expensive hardware is built.