The structure of the reformer comprises a large number of catalyst-filled tubes (where the reforming reactions take place), coupled with burners, that are placed such that the rate of delivery of heat needed to support the endothermic reforming reactions is spatially uniform.
In operation, DRI reformers undergo harsh and variable conditions, which may lead to deformations, and the formation of hot spots and ultimately cracks on the surface of the catalyst-filled tubes [5]. Lowering operating temperatures helps minimize tube material degradation and failure, but lowers efficiency and is thus economically unfavorable. Motivated by economic and safety concerns related to reformer operation, this work introduces a novel model to support operational analysis and optimization of reforming operating conditions, including temperature and production rate.
A substantial body of work is available on modeling conventional steam methane reformers (SMRs) for hydrogen production. However, the differences in operation (including very different process stream compositions and much lower operating pressures in the DRI case) between the two types of reformers mean that SMR models do not readily translate to DRI. Openly available information on reformer model development for DRI applications is quite limited [6-7]. Furthermore, existing models are either lack critical details in terms of geometry and dynamics to be useful for operational optimization, or are overly complex and thus not practical for this purpose [8-10]. These facts motivate the development of a first-principles model that captures features such as radial temperature gradients, tube thermal expansion, and reformer geometry, that are all critical for probing the impact of process conditions on the material of the tubes, and preserving their service.
A bottom-fired upward co-current DRI reformer is modeled, reflecting the configuration used in a typical industrial process and is represented as an ensemble of tube and burner sub-models.
The catalyst-filled tube sub-model considers the relevant length and time scales of the wall and the flow channel inside (as a 2D domain), and the catalyst particles (as a 1D domain of a single particle). The tube-side also considers the thermal expansion effects caused by temperature fluctuations by predicting relative length changes to the tube length. The system of partial differential algebraic equations (PDAEs) is discretized and solved using finite difference numerical schemes. The tube model is calibrated and validated using industrial data, including time-resolved measurements of the process gas temperature at multiple locations along the tube centerline and overall tube thermal expansion, that are provided by a unique set of sensors. A new sensor discrimination algorithm is proposed, to identify sensors that are defective or whose outputs are likely biased due to the harsh operating environment. The tube outlet conditions (i.e., temperature, pressure, composition) show a good agreement between the model and the reference plant data. Particularly, the composition predictions are within 0.5 mole percent of the measured values. Likewise, the model accurately predicts thermal expansion [7].
The burner model is developed to simulate combustion and radiative heat transfer phenomena, then coupled with the tube model. The coupled reformer model can be extended to configurations involving multiple tubes and burners, to capture the spatial effect from reformer specifications. Combustion reaction kinetics are incorporated, which allows for a more accurate prediction of the composition profile along the length domain than in previous works that assume a flame height [6,9,11]. Radiative heat transfer between the burner-side and the tube-side is calculated using the Hottel zone method [12].
The findings of the current study will serve as a foundation for optimizing the plant operation, by developing dynamic operating strategies to minimize thermal stress during mode transitions including startup and shutdown procedures. Optimization efforts will include determining ramp rates and times that mitigate excessive thermal stress, as well as designing fuel flow profiles that maintain safe temperature conditions while ensuring the process gas meets required outlet properties.
References:
[1] IEA (2020). Iron and Steel Technology Roadmap.
[2] DOE (2022). Industrial Decarbonization Roadmap.
[3] IPCC (2006). 2006 IPCC Guidelines for National Greenhouse Gas Inventories.
[4] World Steel Association (2024). World Steel in Figures 2024.
[5] Swaminathan et al. (2008). Failure analysis and remaining life assessment of service exposed primary reformer heat tubes. Eng. Fail. Anal., 15, 4, 311-331.
[6] Latham (2008). Mathematical modelling of an industrial steam methane reformer. M.Sc. Thesis, Queen’s University.
[7] Jeong et al. (2024). An industrially-validated tube model for steam methane reformers used in direct reduced iron production. Int. J. Hydrog. Energy, 91, 1232-1244.
[8] Farhadi et al. (2003). Modelling and simulation of syngas unit in large scale direct reduction plant. Ironmak. Steelmak., 30, 1, 18-24.
[9] Farhadi et al. (2005). Radiative models for the furnace side of a bottom-fired reformer. Appl. Therm. Eng., 25, 14-15, 2398-2411.
[10] Parvathaneni and Andrade (2024). CFD modeling of the industrial-scale bottom-fired direct reduced iron reforming process. Ind. Eng. Chem. Res., 63, 11881-11891.
[11] Quirino et al. (2020). Modeling and simulation of an industrial top-fired methane steam reforming unit. Ind. Eng. Chem. Res., 59, 11250-11264.
[12] Hottel and Sarofim. (1967). Radiative Transfer. McGraw-Hill.