The potential for automotive fuel cells, the desire to maintain combustion stability in supersonic aircraft engines, and the possibility of enhancing lean-premixed flame stability has increased interest in reforming of hydrocarbon fuels for production of hydrogen.  Fuel reforming uses a fuel rich mixture to convert a hydrocarbon (such as straight gasoline) or oxygenated fuel (such as methanol) into H2 and CO or CO2 through some combination of partial oxidation and steam reforming reactions.  Although fuel-reforming technology has existed in the petrochemical industry for many years, petrochemical reformers have been designed for large steady-state systems. With the recent interest in fuel cells for hybrid electric vehicle applications and in hydrogen fuel for high-speed civil transport, the need has arisen for smaller fuel reformers to produce hydrogen both with rapid response and over wide operating windows. At the University of Maryland at College Park, we are developing a program to investigate the coupling of heterogeneous surface reactions and flow fields in H2 reactors.  Our goals are to determine optimal design strategies, both in terms of catalyst structure and surface geometry, for developing fast-response and high-performance fuel reformers in transportation applications.
 
The demands on a fuel reformer will in part depend on the application for which H2 is being generated.  For gas turbine combustion stability enhancement, the reformer will need to avoid excessive temperature rises.  To this end, particular reformer designs will not need to worry about exhaust concentrations of CO and CO2.  A schematic is shown here or a reformer for a gas turbine combustor

On the other hand, reformers used for fuel cell applications, particularly proton-exchange membrane applications, will be highly concerned about minimizing CO in the exhaust (as CO in concentrations greater than 20 ppm is a poison for anode catalysts in PEM fuel cells).   Thus, a reformer for fuel cells will incorporate a downstream water-gas shift reactors which uses additional H2O to react with CO to form more H2 and CO2.  The water-gas shift reactor, which is an area of current research, will also increase the exit H2 mole fraction, which is important for maintaining high fuel cell efficiencies.  A schematic of a staged fuel cell reformer is shown here.

The reactions involved in conversion of hydrocarbon fuels to H2 are three-fold: partial oxidation, steam reforming, and the water-gas shift reaction.  These reactions are summarized in Appendix A for n-heptane.  Designers of fuel reformers will choose some combination of partial oxidation and steam reforming depending upon the availability of steam and heat for the specific application.  For applications with a substantial waste heat and the need for high H2 exit concentrations, steam reforming should be the primary reaction.  However, the indirect heating of the catalyst required for steam reforming will cause the reactor to have slow response times during transient operating conditions such as start-up (unless transient manipulation of H2O/air split is used).  In applications where fast response and limited reactor size is critical, the highly exothermic partial oxidation reactions are beneficial because the heat improves reactor response times and fuel conversion rates.  However, some steam reforming must offset exothermic oxidation reactions to keep reactor temperatures within substrate and catalyst temperature limits.   Some recent development efforts have used H2O/air splits such that the reformer will operate in a so-called autothermal mode, where the heat required and the heat generated are nearly equal.   The operating temperatures in autothermal mode tend to be high enough that there is still a substantial time for heating during start-up, and thus there is a need to develop reactors which provide more rapid start-up.  Our work at the University of Maryland plans to investigate these issues by investigating reforming catalyst additives and alternative surface geometries for improving reactor response and performance at low inlet temperatures.
 

The above plot shows the operating bounds of a fuel reformer operating on a kerosene-like fuel.  Higher H2O inlet concentrations cause the reactor temperature to drop below the limit where catalysts can be expected to be active (around 600 K) and lower H2O inlet concentrations cause the reactor temperature to rise above catalyst temperature limits.  For gas turbine applications where H2 production can improve combustor stability and lower emissions, exit mole fractions of H2 do not need to be high and a designer would favor a lower H2O/air split.  With stoichiometries for H2 and CO production, an H2O mass < 20% of the fuel mass can provide H2 exit mole fractions as high as 33%.  Such H2 exit concentrations can enhance downstream flame stability both for lean and high-speed combustor applications.   For fuel cell applications, higher H2O/air-split would be desirable to produce higher H2 exit concentrations with stoichiometries for H2 and CO2, since even low CO concentrations will poison fuel cell anodes.  For such operating conditions, H2 exit mole fractions above 50% can be achieved before reactor temperatures begin to drop below expected catalyst light-off temperatures.

Our research is looking at the basic performance of reforming catalysts in sub-scale reactors.  The experimental observations will indicate the viability of achieving rapid start-up and minimization of reactor volume and expense.  The study will not investigate some systems integration issues such as managing fuel sulfur content, but results will lay the foundation for which to address systems issues in a more comprehensive design effort.  The study will look at how catalysts, such as nickel-based catalysts with additives to improve light-off and avoid surface carbon build-up, for fuel reforming perform in a simple flow configuration.  A parallel flat-plate reactor will investigate the performance of selected catalysts, which have demonstrated good thermal stability in preliminary oven cycling tests, as a function of inlet temperature, velocity, and H2O/air split.  The experimental results from the performance mapping tests will then be used to validate reduced chemistry models for the catalytic destruction of heavier hydrocarbons.  This study will provide valuable information for actual design of integrated catalytic reforming reactors that can be developed for gas turbine and/or fuel cell applications.

The objectives of our research in fuel reforming center on developing a fundamental catalyst performance data-base for single-component alkane fuels and validating a reduced chemical mechanisms for designing actual reformers with low residence times and high fuel conversion.  Experiments will be used to examine catalyst/substrate compositions and surface mass transfer enhancements for achieving the high fuel conversion and catalyst performance with actual multi-component fuels.  We are seeking to map out the operating conditions of selected catalysts where reactors can achieve high conversion, negligible carbon surface deposition, and thermal stability. The experimentally determined catalyst operating will demonstrate the feasibility of fuel reforming for different applications and the conditions necessary to ensure reliable H2 production from large hydrocarbons.   The study will provide information on the possibility for fuel reformers as a means for stabilizing low temperature flames, for recycling heat in ultra-low emissions power cycles, and for providing H2 for fuel cells.  

For more information, send e-mail to gsjackso@eng.umd.edu
Back to Reacting Flow Laboratory's home page

This appendix reviews global reactions in fuel reformers.  The reaction products are shown as a function of the H2 selectivity (i.e., hH2).  Heats of reaction given below are calculated from heats of formation at 298 K (with liquid values where appropriate) and are given as a function of hH2.  It should be noted that the heats of reaction and thus temperature rise across the reactor can depend strongly on hH2 when inlet stoichiometries are set for H2 and CO formation.  On the other hand, when inlet stoichiometries are set for H2 and CO2 formation, reactor temperature rises are only a weak function of hH2.

As shown in the global chemical reactions below, partial oxidation reactions are exothermic and can be utilized to keep catalyst temperatures high enough to sustain reactions.  When stoichiometries are set to produce H2 and CO care must be taken not to have sudden temperature increases which will favor the formation of carbon and H2O rather than H2.  On the other hand, for H2 and CO2 stoichiometries, heat release rates can be quite high and temperatures are difficult to control without the addition of H2O in the inlet. With air as the oxidizer, pure partial oxidation can only produce up to 0.28 mole fraction H2 in the exit stream with C7H16 (gasoline-like) feeds and no solid carbon formation.

Steam reforming reactions are highly endothermic but H2 exit mole fractions with steam reforming are much higher, up to 0.75 for C7H16.  Higher exit H2 concentrations hoowever are only achievable with large amounts of heat input per mole of reactant.  Required indirect heating of the reactor implies slow response and poor start-up.  This has led some fuel reformer developers to pursue compromises between steam reforming and partial oxidation for applications where response times are critical.  The compromises involve running the reactor with inputs of steam and air in a mode where the net reaction is nearly “autothermal”.

Global partial oxidation reactions generate substantial heat and risk the formation of H2O and CO.

 Stoichiometry for H2 and CO
 C7H16   + 3.5 O2   =>  (7hH2)CO + (7-7hH2)C(s) + (8hH2) H2 + (8-8hH2) H2O
       DHR,298K = (-2062.4+1582.4hH2)kJ/mole
 Stoichiometry for H2 and CO2 (hH2 > 1/8)
 C7H16    + 7.0 O2  => (8-8hH2)CO + (8hH2-1) CO2 + (8hH2) H2 + (8-8hH2) H2O
       DHR,298K = (-2553.2+23.1hH2)kJ/mole

Global steam reforming reactions require substantial heat input but produces primarily CO2 and H2.

 Stoichiometry for H2 and CO2
 C7H16 + 7.0 H2O  => (7-7hH2)CO + (22hH2-15) C(s) + (15hH2) H2 + (15-15hH2) H2O
        DHR,298K = (2062.4-5061.6hH2)kJ/mole
 Stoichiometry for H2 and CO2 (hH2 > 15/22)
 C7H16 + 14.0 H2O  => (22-22hH2)CO + (22hH2-15) CO2 + (22hH2) H2 + (22-22hH2) H2O
       DHR,298K = (1408.3+63.5hH2)kJ/mole

Water-gas shift reaction converts CO and H2O to CO2 and H2 with a small amount of heat release.

CO + H2O => CO2 + H2     DHR,298K = +2.9 kJ/mole