WO2006009453A1

WO2006009453A1 - Catalyst packing, a structured fixed bed reactor and use

Info

Publication number: WO2006009453A1
Application number: PCT/NO2004/000223
Authority: WO
Inventors: Javier PERÉZ-RAMÍREZ
Original assignee: Yara International ASA
Current assignee: Yara International ASA
Priority date: 2004-07-19
Filing date: 2004-07-19
Publication date: 2006-01-26
Anticipated expiration: 2007-01-19

Abstract

A structured catalyst packing comprising one or more catalyst elements consisting of a horizontal layer of individual particles closely packed. Each catalyst particle has one or more parallel flow channels arranged vertically. The layers could be shifted compared to each other or the size and/or location of the flow channels in each layer could be changed allowing for both vertical and radial flow through the packing. The catalyst particles could be cylindrical, hexagonal or square with typically 2-7 flow channels for particles up to 10 mm diameter. Cubic or hexagonal packing could be used for the cylindrical particles. Such a structured catalytic reactor is preferably used for N2O decomposition or NH3 oxidation.

Description

CATALYST PACKING, A STRUCTURED FIXED BED REACTOR AND USE

The invention concerns a structured catalyst packing, a fixed bed reactor for catalytic applications and use of the catalyst.

Packed-bed reactors have been traditionally used in a number of catalytic applications. In this reactor type, the catalyst particles are stacked in a random orientation. Two major drawbacks are identified in conventional packed-bed reactors: high-pressure drop and low catalyst effectiveness. More efficient structured reactors have been developed, including the monolith reactor, the lateral-flow reactor, the parallel-passage reactor, the cross-flow reactor, and the bead-string reactor. As common aspects, these reactors have a low- pressure drop and dust resistance compared to traditional randomly packed reactors. However, they show a serious disadvantage: a low ratio of active catalyst volume to total reactor volume. In particular, for the widely used monolithic reactor, one should take into account some extra difficulties limiting its practical application: relatively complicated manufacture (either by extrusion or deposition of a catalyst layer on an inert support), poor mass and heat transfer from the bulk flow to the catalyst surface, poor radial heat transfer, and absence of radial mass transfer.

The bead-string reactor is characterised by the fact that the catalyst material is attached to parallel wires ordered in a reactor. In practice, the reactor elements of a bead string reactor are manufactured by stringing commercially available hollow extrusions or annular pellets on a wire or rod; or by extruding a catalyst support material around a wire or a rod. Such a reactor is described in Dutch patent application 9000454. This type of reactor has high porosity, i.e. a low catalyst amount-to-reactor volume ratio.

From International patent application WO 9633017 it is known structured reactor packings and a method of manufacturing the same. The catalyst element consists of parallel rods of catalyst-containing material, which material has been provided around a rod-shaped support through pressing in a die. The invention further relates to structured catalyst packing consisting of elements manufactured according to this method. The individual catalyst elements could be arranged relative to each other in such a manner that fluid particles cannot move in a straight line through the packing. Also this reactor type has a very high porosity. International patent application WO 9948604 describes a structured reactor with a framework that divides the reactor into a plurality of elongated cells or chambers within the reactor of different shapes. The framework may be porous or non-porous. The cells are filled with catalyst particles. This configuration only allows vertical flow through the reactor.

The object of the invention is to obtain a reactor and catalyst packing with low-pressure drop and high catalyst volume to reactor volume ratio. Another object is to find catalyst packings suitable for NH₃ oxidation and N₂O decomposition.

These and other objects of the invention are obtained with the reactor and catalyst as described below, and the invention is further defined and characterised by the accompanying patent claims. The invention will be further illustrated with reference to the figures 1-6, where

Figure 1 shows different catalyst shapes applied.

Figure 2A shows closely packed pellets of hexagonal shape and Figure 2B shows a packing of pellets with square shape.

Figure 3A shows line-up cylinders in hexagonal packing, while Figure 3B shows shifted cylinders in hexagonal packing. Figure 3C illustrates line-up cylinders in cubic packing, while shifted cylinders are shown in Figure 3D.

Figure 4 shows bed porosity for different reactor configurations.

Figure 5 shows a basket used in pilot plant tests with random fixed-bed reactor A and the reactor according to the invention C. The middle picture, B, shows a mixed configuration.

Figure 6 shows the influence of the reactor configuration on the bed porosity, pressure drop, and catalytic performance in the high-temperature ammonia oxidation and in direct N₂O decomposition.

The invention thus concerns a structured catalyst packing that comprises one or more catalyst elements consisting of a horizontal layer of individual catalyst particles closely packed where each catalyst particle has one or more parallel flow channels arranged vertically. The layers could be aligned or shifted compared to each other allowing for both vertical and radial flow of gas through the packing. Radial flow could also be obtained by changing the size and location of the channels in each layer. In this way radial flow could be obtained in cubic, hexagonal and cylindrical aligned particles as well. The catalyst particles are preferably cylindrical, hexagonal or square. A catalyst particle up to 10 mm diameter will have 2-7 flow channels. The number of flow channels will be dependent on the particle size. A large amount of flow channels is preferred to increase the surface area, but it should not be more than to still maintain the strength of the particles. The particle diameter is preferably d_p = 3-30 mm.

A preferred configuration is cylindrical catalyst particles arranged in cubic packing. Another preferred configuration is cylindrical catalyst particles or catalyst bodies arranged in hexagonal packing. Preferably the open space between the particles in this packing has an area in the same order as each flow channel in the particle. This allows a very uniform flow. Every second layer of catalyst particles could be shifted.

The invention also concerns a fixed-bed reactor with a catalyst packing as described above. Preferred use of such a reactor is for NH₃ oxidation and N₂O decomposition. For NH₃ oxidation it is preferred to use a catalyst packing with aligned channels in all layers, while for N₂O decomposition it is preferred to use a catalyst packing with shifted channels or a packing where the size and/or location of the channels are changed in each layer to allow both vertical and radial flow through the packing.

The invention described here focuses on the development of a novel type of reactor with structured catalyst packing for heterogeneous catalysis, the so-called modular structured fixed-bed reactor (MOSFIBER).

In essence, MOSFIBER is a fixed-bed reactor with ordered catalyst material of different shapes. This configuration displays the advantages of monolithic reactors (low-pressure drop), but having a high catalyst amount-to-reactor volume ratio, even higher that in traditional fixed-bed reactors with random packing of catalyst particles. Hydrodynamics, kinetics, and transport phenomena can be decoupled in this configuration, which is highly advantageous in catalytic applications.

Figure 1 shows different catalyst shapes applied, in the form of a square, hexagon or cylinder. Several reactor configurations with different geometrical characteristics can be considered by structuring the pellets in different ways. The examples are carried out using equilateral pellets (length and diameter are the same, ca. 10 mm and holes (flow channels) of ca. 2 mm), but any dimension of particles could be used. The number of holes or flow channels of the pellets is variable, and also dependent on the size, but it should have more than one. Compared to a monolith, the main difference is a smaller size of the pellets (denoted as miniliths) and that a conventional pressing process can manufacture them. A monolith requires extrusion when the whole monolith is to be made of the active phase, which is technically challenging. A ceramic monolith (often cordierite) can be wash coated with a thin layer of catalyst. In this case, the catalyst amount is even lower. In addition, it is hardly possible to make an even layer of catalyst and stabilizing it under reaction conditions, i.e. wet oxidizing atmosphere at high temperature.

Figure 2 and 3 shows the arrangement of pellets. Figure 2A shows closely packed pellets of hexagonal shape and Figure 2B shows a packing of pellets with square shape. In the case of cubes and hexagons, the holes in the pellets are aligned so macroscopically the reactor is a monolithic reactor. However, no monolith can be extruded with the flexibility in dimension of pellets (particularly regarding hole size). However, it should also be mentioned that it is possible to use cubes with a different number and/or position of holes, destroying the alignment. This is exemplified with cylinders, as shown in Figure 3, where apart from cubic and hexagonal packing within a layer, full alignment or shifted layers can be considered. This will affect the hydrodynamics and transport phenomena, and depending on the nature of the reaction, the catalytic performance. Figure 3A thus shows line-up cylinders in hexagonal packing, while Figure 3B shows shifted cylinders in hexagonal packing. Figure 3C illustrates line-up cylinders in cubic packing, while shifted cylinders are shown in Figure 3D. By using shifted particles, horizontal flow through the catalyst bed is made possible.

The orientation and packing of the catalyst can be changed as shown in Figure 3, so the reactor characteristics {e.g. porosity) and thus performance can be tuned to resemble anything between a conventional randomly distributed fixed-bed reactor and a monolithic reactor.

Figure 4 shows bed porosity as a function of reactor configuration for different materials. Column 1-4 illustrates catalyst materials according to this invention compared to a random bed in column 5 and monoliths in columns 6-7. The figure shows that the structure in MOSFIBER leads to reactors with low porosities, ranging from 20-40 %, enabling high amounts of catalyst per reactor volume. These values are considerably lower than a random fixed-bed reactor and dramatically differ from the high porosities in monolithic reactors, up to 80 % depending on cell density and wall thickness.

Preliminary calculations showed the high potential of this reactor configuration for short- contact time gas-solid reactions with selectivity issues (e.g. partial oxidation of methane, ammonia oxidation, ethane dehydrogenation), and also in processes limited by the reactor volume, where a high catalyst loading is required to achieve a certain reactant conversion (e.g. N₂O decomposition.) The development was completed with pilot plant tests with MOSFIBER modules and the validation of a reactor model. The activity of the MOSFIBER reactor for in-process N₂O decomposition is up to 40% higher as compared to the fixed- bed reactor configuration (using the same amount of catalyst and shape). Furthermore, increased selectivities are obtained in NH₃ oxidation. The performance of MOSFIBER in these applications is also superior to monolithic reactors.

The invention will be further illustrated with reference to the examples.

Example 1.

Pilot tests were carried out to examine the influence of the reactor configuration on the bed porosity, pressure drop, and catalytic performance in the high-temperature ammonia oxidation and in direct N₂O decomposition.

Figure 5 shows the catalyst basket used in the pilot tests. It was used cylindrical pellets with diameter about 10 mm and 7 flow channels. The catalyst amount (or number of pellets) in the pictures as shown in Figures 5a and 5c is exactly the same (ca. 2 kg), but the reactor volume in Figure 5a is obviously higher (due to the higher porosity). The top view in Figure 5b nicely illustrates the degree of structure in MOSFIBER as compared to fixed-bed configurations normally used in industrial gas-solid applications.

Ammonia oxidation was tested under the following conditions: 10.5 vol. % NH₃ in air, 900⁰C, 5 bara, GHSV = 200,000 h^'1. N₂O decomposition was tested installing the catalyst beneath Pt-Rh gauzes during ammonia oxidation at the following conditions: 1500 ppm N₂O, 7 vol.% O₂, 10 vol.% NO, and 13 vol.% H₂O in N₂, 87O⁰C, 5 bara, and GHSV = 200,000 h^'1. In all cases, the catalyst was CO₂AIOVCeO₂ containing 1.2 wt.% Co.

Figure 6 shows the influence of the reactor configuration on the bed porosity, pressure drop, and catalytic performance in the high-temperature ammonia oxidation and in direct N₂O decomposition. Experiments were carried out both for a random fixed-bed reactor and MOSFIBER with shifted cylindrical pellets and hexagonal packing and with cubic pellets. The results are also compared with monoliths. These have a limited amount of catalytic material, as the catalytic material is applied as a wash coating to the ceramic structure. For this purpose, the volume of the monolithic reactor was the same as for MOSFIBER. Accordingly, all the catalysts are compared in terms of reactor volume.

As can be seen from the figure, the porosity in MOSFIBER configurations is considerably lower as compared to random fixed-bed reactor and monolithic reactor. The monolith used in this example was that of 200 cpsi in Figure 4. In spite of the extremely low porosity of MOSFIBER, the relative pressure drop is very similar to that of the monolith and 5 times lower as compared to a random fixed-bed reactor. This is due to the presence of structure. The performance of MOSFIBER in both catalytic applications investigated is superior to the other reactor configurations. Again, the poor performance of the monolithic reactor is a consequence of the limited amount of catalyst per reactor volume.

Two MOSFIBER configurations have been compared in this example: shifted cylindrical pellets in hexagonal packing and cubic pellets. In both cases, the volume of holes per pellet is the same in order to obtain a proper comparison. The N₂O conversion over the cylindrical configuration is slightly higher than over the cubic configuration, while the opposite holds in ammonia oxidation. This nicely illustrates that the mechanism of each reaction and the transport phenomena associated in the reactor scale determine the optimal configuration. The rate of direct N₂O decomposition is significantly lower than that of ammonia oxidation and apparently is favoured by the radial dispersion component introduced by shifting the catalyst pellets. Ammonia oxidation is a very fast, selectivity- controlled reactor and aligned channels seem to be preferred.

MOSFIBER thus combines a tunable and low porosity (high W_cat/V_reacto_r) and a low pressure drop (in the same order as monoliths). Which MOSFIBER configuration to be applied depends on the intrinsic characteristics of the reaction and reactor space available in retrospective applications.

Claims

1. A structured catalyst packing, characterised in that it comprises one or more catalyst elements consisting of a horizontal layer of individual catalyst particles closely packed, where each catalyst particle has one or more parallel flow channels arranged vertically.

2. Catalyst packing according to claim 1 , characterised in that the layers are shifted compared to each other allowing for both vertical and radial flow of gas through the packing.

3. Catalyst packing according to claim 1 , characterised in that the size and/or location of channels is changed in the different layers to obtain both vertical and radial flow through the packing.

4. Catalyst packing according to claim 1 , characterised in that the catalyst particles are cylindrical, hexagonal or square.

5. Catalyst packing according to claim 4, characterised in that the catalyst particles have particle size d_oat = 3- 30 mm.

6. Catalyst packing according to claim 1 , characterised in that the catalyst particles are cylindrical and arranged in cubic packing.

7. Catalyst packing according to claim 1 , c h a r a ct e r i s e d i n t h at the catalyst particles are cylindrical and arranged in hexagonal packing.

8. Catalyst packing according to claim 6, characterised in that the space formed between the particles in the packing has an area in the same order as each flow channel in the particles.

9. Catalyst packing according to claim 1 , characterised in that every second layer is shifted.

10. Fixed-bed reactor, characterised in that it comprises a catalyst packing according to any of the claims 1 -9.

11. Use of a fixed bed reactor according to claim 10, for NH₃ oxidation.

12. Use according to claim 11 , where the catalyst packing has aligned channels in all layers.

13. Use of a fixed bed reactor according to claim 10, for N₂O decomposition.

14. Use of a fixed bed reactor according to claim 13, where the catalyst packing has shifted channels or a packing where the size and/or location of the channels are changed in each layer to allow both vertical and radial flow through the packing.