Advances in multiphysics modelling and scale-up pathways for microwave-assisted pyrolysis in bioenergy applications

Achieving a sustainable energy future requires efficient renewable conversion pathways, with biomass emerging as a promising alternative. Microwave-assisted pyrolysis (MAP) has attracted growing interest due to its high energy efficiency and product yields, yet its complex interplay of electromagnetic, thermal, and chemical processes demands advanced modelling for reliable design and optimisation. Despite the use of diverse software tools to study MAP, systematic evaluations of their capabilities remain scarce. This review critically assesses major modelling platforms used in MAP research, including process simulators (e.g., Aspen Plus, Aspen HYSYS), Computational fluid dynamics (CFD)-based solvers (e.g., COMSOL Multiphysics, ANSYS CFX, OpenFOAM), and statistical or machine learning environments (e.g., MATLAB, Design Expert). Their applications are compared in terms of feedstock dependence, operating conditions, and modelling features. Process simulators are particularly effective for flowsheet analysis and techno-economic studies, while CFD tools capture transport phenomena and reactor-scale behaviour with high resolution. Data-driven platforms complement these approaches by enabling optimisation and predictive analytics. Given the complexity of MAP, a modular modelling strategy is recommended, treating stages such as drying, heating, and pyrolysis independently with tailored methods. By consolidating existing knowledge and identifying gaps, this review provides a practical guide for researchers and engineers to select and integrate the most suitable numerical approaches for advancing MAP system development. • Reviews recent developments in multiphysics modelling of microwave-assisted pyrolysis (MAP). • Highlights the lack of generalised reaction kinetics across diverse feedstocks as a key challenge in MAP modelling. • Identifies material properties, reactor design, and operating conditions as major barriers to upscaling and commercialisation. • Proposes CFD–machine learning integration as a promising route for optimised MAP reactor design and scale-up.