Biomass pyrolysis process simulation for design and assessment: A critical, decision-oriented review of aspen plus workflows and complementary tools

Abstract

Biomass-pyrolysis process models increasingly support flowsheet design, heat integration, optimization, and techno-economic or environmental assessments. However, the reliability of their conclusions depends on whether feedstock heterogeneity, conversion behavior, condensable products, and validation evidence are represented adequately for the intended engineering decision. This review critically evaluates Aspen Plus workflows based on fixed-yield, thermodynamic-equilibrium, and kinetic formulations, while explicitly retaining the feedstock, technology, and software context required for defensible model selection. A curated 2022–2026 corpus comprises 51 model implementations from 50 unique publications. Within this corpus, equilibrium, kinetic, and fixed-yield/product-distribution approaches account for 31.4%, 31.4%, and 37.3%, respectively, compared with 63.0%, 14.0%, and 23.0% in the 2010–2021 survey. A four-level validation audit shows that 17 of the 51 implementations provide broad independent quantitative evidence, whereas 21 lack independent validation of the primary product-generating model, including instances where outlet yields are prescribed or calibrated without an external test. Because reported validation metrics vary substantially in definition, output, feedstock, and operating domain, a pooled error comparison across model classes is not scientifically defensible. Fixed-yield models support condition-matched process integration but generally provide limited independent predictive capability outside the source-data domain. Equilibrium models provide useful thermodynamic and gas-phase benchmarks but should not generally be treated as stand-alone predictors of condensable bio-oil. Among the reviewed implementations, kinetic models provide the strongest evidence base for operating-condition prediction, provided that parameter provenance, reactor and transport assumptions, applicability limits, and independent validation are demonstrated. The resulting decision framework links simulation objectives and available evidence to model fidelity, software role, validation requirements, and defensible claim boundaries. Priority needs include transferable kinetic datasets, traceable feedstock and product representations, explicit uncertainty treatment, and time-resolved validation for dynamic models.