Patent Application
20090042071
The present invention relates to catalytic hydrocarbon fuel reformers for converting a hydrocarbon stream to a gaseous reformate fuel stream comprising hydrogen; more particularly, to fuel reformers having heat exchange capability; and most particularly to a multi-tube endothermic fuel reformer having a high surface area metal substrate attached to both the insides and the outsides of the reforming tubes to augment heat transfer to the reforming catalyst. The present invention is useful for providing reformate as a fuel to a fuel cell, especially a solid oxide fuel cell, or to an internal combustion engine.
A catalytic hydrocarbon fuel reformer containing a fuel reforming catalytic reactor converts a fuel stream comprising, for example, natural gas, light distillates, methanol, propane, naphtha, kerosene, gasoline, diesel fuel, or combinations thereof, and air, into a hydrogen-rich reformate fuel stream comprising a gaseous blend of hydrogen, carbon monoxide, and nitrogen (ignoring trace components). In a typical reforming process, the raw hydrocarbon is percolated with oxygen in the form of air through a catalyst bed or beds contained within one or more reactor tubes mounted in a reformer vessel. The catalytic conversion process is typically carried out at elevated catalyst temperatures in the range of about 700° C. to about 1000° C.
The produced hydrogen-rich reformate stream may be used, for example, as the fuel gas stream feeding the anode of an electrochemical fuel cell. Reformate is particularly well suited to fueling a solid-oxide fuel cell (SOFC) system because a purification step for removal of carbon monoxide is not required as is the case for a proton exchange membrane (PEM) fuel cell system.
The reformate stream may also be used to fuel a spark-ignited (SI) engine, either alone or in combination with gasoline. Hydrogen-fueled vehicles are of interest as low-emissions vehicles because hydrogen as a fuel or a fuel additive can significantly reduce air pollution and can be produced from a variety of fuels. As a gasoline additive, controlled amounts of supplemental hydrogen fuel may allow conventional gasoline-fueled internal combustion engines to reach nearly zero emissions levels.
Prior art fuel reformers such as catalytic partial oxidation (CPOx) reformers have limitations in maximum obtainable reforming efficiency. Prior art CPOx endothermic reforming delivers maximum reforming efficiencies of between about 78% and 82% for DF2 diesel fuel, gasoline, or JP8 jet fuel, and up to about 84% for methane and natural gas, at usual reactant preheat temperatures of about 300° C. Only if the efficiency of an associated fuel cell stack is greater than 80% can such a system deliver overall system efficiencies of greater than about 30-35%. On the other hand, endothermic catalytic reformers using heat exchangers to boost the reforming temperature can deliver efficiencies of about 120%, which allow for fuel utilizations in the stack of 40-60% and overall system efficiencies greater than 35%. Such efficiencies require endothermic heat exchange construction having a) very high heat transfer capability; b) a large surface area for washcoat admission; c) high durability and long lifetime of the washcoat substrate; and d) high durability and long lifetime of the catalytic washcoat on the substrate.
Prior art endothermic reactors typically are formed of planar elements, some of which are known to have a plurality of small channels (microchannels) formed in the planar surfaces to increase surface area. Such reactors have shown maldistribution of heat and poor washcoat adhesion. During coating of the washcoat slurry onto the reactor surfaces, only small amounts of washcoat are picked up; the washcoat cannot be controlled as to where and how much of it will be deposited; and thermal differential expansion (TDE) of the base metal of the reactor and of the washcoat can cause the washcoat to spall off within a few minutes of the start of operation. To reduce the TDE problem, some prior art reactors are formed of expensive high temperature alloys, such as Haynes alloys. Prior art microchannel planar reactors are complex and expensive to design and fabricate; cannot be repaired after failure; and are susceptible to thermal imbalance and washcoat spalling.
What is needed in the art is a catalytic reforming reactor having a) very high heat transfer capability; b) a large surface area for washcoat admission; c) high durability and long lifetime of the washcoat substrate; and d) high durability and long lifetime of the catalytic washcoat on the substrate.
It is a primary object of the invention to provide a durable, less expensive endothermic hydrocarbon fuel reformer capable of substantially increasing fuel efficiencies.
A catalytic reformer assembly includes a reactor comprising a plurality of parallel tubes arranged within a tubular housing. A metal substrate, formed as corrugated sheet metal, rolled wire mesh, wire gauze, or preferably a metal foam lattice, is attached as by brazing to the parallel tubes on both their inside and outside surfaces. A catalytic washcoat is applied to the metal substrate within the tubes, defining thereby the interiors of the tubes as catalytic reforming reactors. The endothermic reforming reactions within the tubes are supported by heat from hot combustor exhaust gas flowing around the tubes in contact with the augmenting heat transfer metal substrates outside the tubes. Radial temperature gradients are small because of excellent heat transfer across the tube walls, resulting in excellent mechanical stability of the washcoat on the metal substrate. Preferably, the tubes are formed of Inconel 625 and the metal substrate is formed of Fecralloy®, a high temperature alloy having excellent thermal conductivity and oxidation resistance.
The present invention will now be described, by way of example, with reference to the accompanying drawings, in which:
Referring to
In operation, incoming fuel 26 is vaporized by heat exchanged from combustor exhaust 22 either before or after being mixed with air, steam, and recycled anode tail gas 28. The flows of air, steam, and recycled anode tail gas are individually controlled and variable, as is known in the prior art of endothermic reforming. The resulting fuel stream 30, typically at a temperature of about 150° C., is fed to endothermic reactor 14 wherein reformate 32 is produced and supplied to stack 12 at a temperature of about 900° C. Stack 12 combines hydrogen and carbon monoxide in reformate 32 with oxygen from air (not shown) supplied to stack 12 to produce electricity, anode tail gas 18, and cathode tail gas 20, each at a temperature of about 750° C. Anode tail gas is rich in unreacted hydrogen and carbon monoxide, as well as water and carbon dioxide formed in the electrochemical reaction which produced electricity.
A first portion of anode tail gas 18 is supplied (not shown) to assist in forming fuel supply mixture 28.
Endothermic reforming conditions within reactor 14 are achieved by augmented heat transfer as described below. Combustor 16 receives a second portion of anode tail gas 18 and cathode tail gas 20 which form a combustible mixture that is ignited within combustor 16 to generate a hot exhaust 22 at a temperature of about 1150° C. The combustible mixture within combustor 16 may be enriched as desired by the addition of gaseous or pre-vaporized hydrocarbon fuel 34.
The generic use of combustor heat to assist in pre-heating of reformer reactants and the reforming reactor is known in the art. The improvement defined by the present invention is the arrangement, construction, and operation of the endothermic reformer reactor 14 to a) minimize radial thermal gradients between the inside and outside of the individual reactor tubes in order to minimize thermal stresses on the washcoat and thereby eliminate spalling of the washcoat from its substrate within the tubes; b) assure uniform distribution and high surface area of the washcoat on its substrate within the tubes; and c) provide reforming efficiencies in excess of 120%.
Referring now to
Tubes 42 are formed of a high temperature alloy such as Inconel 625. In one aspect of the invention, tubes 42 are between about 4 inches and about 10 inches in overall length and between about ¼ inch and about 2 inches in outside diameter, depending upon requirements for reforming efficiency and electric power from the fuel cell stack. A first heat-transferring substrate 48 formed of corrugated rolled sheet metal, rolled wire mesh, wire gauze, or preferably a metal foam lattice surrounds tubes 42 in the inter-tube space 50 therebetween and is bonded to the tubes as by brazing or welding to provide a low-resistance heat transfer path from the substrate to the tubes. First substrate 48 is arranged such that a free path exists within space 50 for axial flow of combustor exhaust gases 22 as described below. Brazing with a high level of contact between the first substrate and the tubes is achieved by spray coating the bare tubes with braze material, installing the fist substrate in appropriate contact with the tubes, and then baking the assembly at an appropriate temperature to cause the braze material to liquefy and bond to the first substrate and the tubes at a large plurality of points.
A second substrate 52 is installed within each of tubes 42 and preferably is similarly formed and bonded to the tubes in a fashion similar to the bonding of first substrate 48. The baking of the second substrates preferably is carried out with the baking of the first substrate to yield a reformer reactor subassembly 14a , as shown in
Preferably, first and second substrates 48,52 are formed of the same material. Referring now to
A challenge in the use of metal foams, as in the present application, is increased oxidation promoted by the high specific surface area and high temperatures, which can lead to catastrophic failure of the material. The presently preferred material for forming a foam in accordance with the present invention is Fecralloy® alloy, available from Goodfellow Cambridge Limited, Huntingdon, England. This material is an alloy of iron, chromium, aluminum, yttrium, and zirconium.
After subassembly 14a is formed, second substrate 52 within the tubes is coated with a reforming catalyst material (not shown) in known fashion, preferably by liquid coating of a washcoat onto second substrate 52 followed by drying and sintering to adhere the catalyst material to the entire exposed surface area of substrate 52.
An added benefit of using Fecralloy® for the substrate, both inside and outside of the tubes, is that it has very high thermal conductivity, resulting in high heat transfer from gas in space 50 through the tube walls and into the reactor space within the tubes. Thus, there is a very low radial thermal gradient across the tube walls. As noted above, a serious problem in the prior art is that the catalyst material is easily spalled from the prior art substrate surfaces due to high thermal gradients imposed by low thermal conductivity of the substrate materials and large planar surface areas of the substrates.
Referring now to
Referring now to
In operation, as shown in
Referring to
It is seen that the gross efficiency of the reformer with methane (curve C) and without methane (curve D) is nearly 140%, clearly superior to prior art endothermic reformers that demonstrate only about 120% gross efficiency. The respective net efficiencies are shown in curves A and B. The mole percentages of components of the resulting reformate are shown for carbon monoxide (curve F), hydrogen (curve E), water (curve G), carbon dioxide (curve H), and methane (curve 1).
The benefits of an improved endothermic reformer over prior art endothermic reformers may be summarized as follows:
1. The reformer may be plumbed for either co-flow or counter-flow of combustor exhaust, for highest efficiency in any given application.
2. Small radial temperature gradients promote long working life for the substrates and the catalyst washcoat.
3. Employing relatively few, large reforming tubes reduces cost and simplifies construction.
4. The combustor is external to the reformer, making it simpler to install and repair, and allowing for improvement in the combustor without redesign of the reformer; some prior art reformers include an integral combustor.
5. Use of Inconel 625 for the tubes reduces cost over prior art use of Haynes alloys.
6. The substrates are formed of high conductivity materials such as Inconel or Fecralloy.
7. Washcoat adhesion is much improved on the irregular latticework of the second substrate, with low impetus for thermal spalling.
8. The first and second substrates are readily brazed to the tubes for maximum heat transfer from the combustor exhaust to the fuel mixture.
9. The first and second substrates can be readily removed and replaced as may be required from time to time by melting out the braze.
10. New catalyst can be applied to the second substrate without requiring a new reactor.
11. The components may be either welded or brazed together.
12. The overall assembly is compact, relatively inexpensive, and easy to manufacture.
13. The design can be readily adjusted for different power range requirements simply by adding or removing tubes from the reactor design.
While the invention has been described by reference to certain preferred embodiments, it should be understood that numerous changes may be made within the spirit and scope of the inventive concepts described. Accordingly, it is intended that the invention not be limited to the disclosed embodiments, but that it have the full scope permitted by the language of the following claims.
