Scaling up photoelectrochemical (PEC) devices to commercially viable industrial systems requires a comprehensive understanding of coupled transport phenomena across multiple length scales, which can be enabled by numerical modelling to guide system design and optimization. This is especially important for systems using unconventional substrates like recently developed transparent porous conductive substrates (TPCS) [1]. These novel substrates offer significant increases in electrochemically active surface area but also pose challenges with light scattering and recombination losses. This work presents a holistic, bottom-up multi-scale and multi-physics modelling framework for a PEC reactor featuring porous photoelectrodes made by depositing photocatalysts on TPCS fibers. The system is evaluated for hydrogen evolution combined with glycerol oxidation (an industrial waste byproduct), promising a waste-valorisation pathway.

The presented modelling framework uniquely bridges microscale optics with macro-scale reactor engineering. First, the interaction of individual multilayered microfibers with the incident radiation was resolved and subsequently used to compute the radiation characteristics of a thick homogeneous fibrous medium. A Monte Carlo ray tracing method was developed to simulate light propagation through the medium, accounting for fiber orientations and multiple scattering. The local photon absorption profile along the photoelectrode thickness was computed precisely and coupled with charge transport and heat transfer physics to determine local photocurrent and heat generation rates.

To capture device-level behaviour, localized data was integrated into a reactor-level homogenized continuum model. This macro-scale model simultaneously simulates light transport, fluid flow, heat transfer, charge transport, and species transport. After verifying predictions against experimental measurements on equivalent systems, the model was used to quantify performance metrics such as photocurrent density, temperature, and local overpotentials under varying solar concentrations and flow rates to identify limiting regimes and optimization pathways. Additionally, model predictions with and without glycerol were compared.

Results demonstrated that performance was limited by significant ohmic losses at large current densities and non-uniform light absorption inside the porous sample. As solar concentration C_sun increased, the performance gap between the ideal uniform case and real cases widened further due to escalating optical and ohmic losses. Convective heat transfer from fluid flow maintained reactor temperatures below 45°C, even under high irradiation (C_sun = 50). Concentration gradients confirmed successful glycerol oxidation and hydrogen evolution, while the impact of different material and operating parameters on the J-V curves was quantified.

This model offers a powerful framework for the rational design of next-generation PEC reactors incorporating porous photoelectrodes, accelerating their transition to real-world solar-fuel deployment.