Ebullated bed reactors for heavy oil upgrading: A comprehensive review of technology, hydrodynamics, and computational modeling
Issued Date
2026-02-15
Resource Type
ISSN
00162361
Scopus ID
2-s2.0-105016995271
Journal Title
Fuel
Volume
406
Rights Holder(s)
SCOPUS
Bibliographic Citation
Fuel Vol.406 (2026)
Suggested Citation
Lumphoon K.N., Nuntang S., Nimmanterdwong P., Manatura K., Piemjaiswang R., Chalermsinsuwan B. Ebullated bed reactors for heavy oil upgrading: A comprehensive review of technology, hydrodynamics, and computational modeling. Fuel Vol.406 (2026). doi:10.1016/j.fuel.2025.136977 Retrieved from: https://repository.li.mahidol.ac.th/handle/123456789/114640
Title
Ebullated bed reactors for heavy oil upgrading: A comprehensive review of technology, hydrodynamics, and computational modeling
Corresponding Author(s)
Other Contributor(s)
Abstract
Despite decades of industrial application, significant knowledge gaps remain between ebullated bed reactor (EBR) laboratory studies and commercial-scale implementation, particularly regarding computational fluid dynamics (CFD) validation under industrial conditions. As conventional crude reserves decline, the energy industry increasingly relies on heavy oils comprising 70 % of global reserves, necessitating advanced EBR processing technologies operating at high pressures (10 − 20 MPa) and temperatures (380 − 440 °C). This comprehensive review provides the first systematic analysis linking CFD evolution with industrial implementation challenges, revealing critical gaps that limit current predictive capabilities. Novel insights emerge from analyzing complex multiphase hydrodynamics, while persistent challenges include computational limitations and inadequate mesophase formation modeling. The review critically evaluates CFD methodologies from Eulerian-Eulerian (E-E) to Eulerian-Lagrangian (E-L) frameworks, identifying fundamental limitations in drag closure models under high gas holdup conditions. Three primary knowledge gaps are identified: (1) CFD validation limitations with accuracy degrading from ± 8 % at pilot scale to ± 25 % commercially, (2) mesophase formation prediction inadequacies affecting catalyst performance, and (3) machine learning (ML) integration challenges where promising results (R<sup>2</sup> > 0.99 for gas holdup) deteriorate when extrapolating beyond training conditions. Novel multi-scale modeling approaches and hybrid ML-CFD methodologies show promise but require physics-based constraints. Future research priorities include developing validated closure relations for industrial conditions, advancing mesophase modeling through in-situ characterization, and establishing robust ML frameworks for sustainable heavy oil upgrading optimization.