![[ Visit Client Website ]](images/banner.gif)
Introduction
The worldwide hydrogen demand has strongly increased in the past years, mostly for refinery needs in order to match environmental requirements such as desulfurization. This megatrend is expected to maintain for a long term period and raise new challenges associated to the design and operation of large scale hydrogen plants, as of today mostly relying on steam reforming of natural gas [ 1 ].
As a major actor in the field of hydrogen production, and hydrogen plant engineering through its subsidiary LURGI, AIR LIQUIDE has developed advanced in-house simulation tools to understand, predict and optimize the efficiency, the compactness and the reliability of large scale steam methane reformers.
The present paper aims at presenting the SMR3D modeling tool and its application to key aspects of a recent Lurgi reformer design.
SMR3D presentation
ATHENA, the AIR LIQUIDE in-house CFD code developed since the 80's, is a robust industrial tool, which includes several established models dedicated to heat transfers in industrial furnaces, including oxy-combustion [ 2 ]:
- The averaged three-dimensional flow is solved by use of a RANS approach (Reynolds Average Navier-Stokes). The Navier-Stokes equations are computed averaged on time, while turbulent fluctuations are modeled with using the standard k–e model [ 3 ].
- The solver is based on a cell-centered, structured finite volume discretization of Navier-Stokes equations [ 4 ] and on transport equations of passive or active scalars (enthalpy, mass fractions, turbulence quantities of RANS models …), with a SIMPLE-algorithm segregated solution procedure.
- Radiative heat transfer equations are solved by use of a specific ray tracing method (DTRM: Discrete Tracing Radiative Method [ 5 ]), a method which proved to be adapted to complex geometries. The gas radiation properties are taken into account with a WSGG model (Weighted Sum of Grey Gases), with four gray gases tables available for air and oxy- combustion products of different fuel compositions [ 6 ], with or without soot.
- The Magnussen model [ 7 ] is commonly used for industrial furnaces simulations, because it provides several pertinent tuning parameters, and then enables to use coarse mesh suited to such large geometries, and then avoid to fully detailing the burner geometry.
The computation of the flow, heat transfers and chemical endothermic reaction of steam methane reforming inside the catalytic tubes has been carried out using the code developed at Gent University described in [ 8 ][ 9 ]. The tube is described as a 1D plug-flow reactor in the vertical direction and computes the Hougen-Watson reactions set:
The reforming model takes into account the intrinsic catalytic kinetics and the limitation though the diffusion into the porous catalysts, using reaction effectiveness factors, ηr. Although these effectiveness factors could be computed by the gent code, these parameters were considered as constant in order to speed-up the simulations. Here, the effectiveness factors were kept constant and their values were established on the basis of operational data gathered on AIR LIQUIDE operated plants.
The SMR3D solver has been developed in order to take advantage of large computers. Two levels of parallelization (MPI + OpenMP) allow running massively parallel computations involving hundreds of million cells with a reasonable restitution time. Therefore, simulation of full scale large reformer with hundreds of tubes and burners is now at hand.
SMR3D has been validated through confrontation with experimental data gathered on AIR LIQUIDE operated fireboxes.
Results applications
On one hand, the capacity of SMR3D to describe the governing physical phenomena inside the firebox, despite the difficult experimental access, is a clear motivation. For example, it enables
- heat loss characterization and mapping,
- flow arrangement characterization
- hotspots identification and tube lifetime improved management
- tube duty distribution, etc…
- safe process intensification
In this paper, the analysis of the SMR3D simulation results is presented on a reformer case, allowing to get in depth understanding of the coupling effects between the combustion chamber and the reforming tube, with a level of details not reported in the literature to our knowledge.
On the other hand, the continuous improvement of the commercial LURGI reformer is benefitting from SMR3D simulation capabilities. As an illustration, two of the largest AIR LIQUIDE operated plants, recently designed by LURGI, were simulated and the results are compared (~400 tubes each). The HERACLES plant has been commissioned in 2011, whereas the second project is to be started-up in 2013, and could benefit of SMR3D inputs to optimize the tube duty distribution and the flue gas flow arrangement, which yields to significant performance improvements.
Conclusions
SMR3D is the results of 25 years of continuous efforts for rigorous integration of the most appropriate models, while keeping the solver highly efficient. It is now possible to run simulations of full scale large SMR furnace, including combustion, radiation and the coupling of combustion chamber to individually calculated reforming tubes.
The understanding of the key phenomena involved in reformers operation and design has been greatly improved, yielding to further optimization and benefits in terms of efficiency, reliability and compactness.
References
[ 1 ] New York Power Authority, “Hydrogen fact sheet: Hydrogen production – Steam Methane Reforming (SMR)”, www.getenergysmart.org, 2005
[ 3 ] Jones, W.P., Launder, B.E., “The prediction of laminarization with a two equation model of turbulence”, Heat and mass transfer, 15, pp. 310-314, 1972.
[ 4 ] Patankar, S.V., “Numerical Heat Transfer and Fluid Flow”, Hemisphere, 1980.
[ 5 ]
Lockwood,
F.C. and Shah, N.G., “A new radiation solution method for incorporation in
general combustion prediction procedures”, 18th Symposium (International) on
Combustion, The combustion Institute, [ 6 ]
Soufiani, A., Djavdan, E., “A comparison between weighted sum of grey gases
and statistical narrow band radiation models for combustion applications”,
Combustion and Flame, Vol. 97, 1994 [ 7 ] Magnussen, B.F., Hjertager, B.H.,
“On Mathematical Modeling of Turbulent Combustion with Special Emphasis on Soot
Formation and Combustion”, 16th Symp.
(Int.) on Combustion, 1976. [ 8 ]
Xu,
J., Froment, G.F., “Methane Steam Reforming – I Methanation and Water-Gas Shift: Intrinsic kinetics”; AIChE journal; 1989. [ 9 ]
Xu, J., Froment,
G.F., “Methane Steam Reforming – II Diffusional
limitations and reactors simulation”; AIChE journal; 1989.
See more of this Group/Topical: Catalysis and Reaction Engineering Division