Oxygen and nitrogen injection for increasing ammonia production

Abstract (English, WO 1998045211 A1)

Ammonia production is increased in an ammonia plant having a primary reformer, a secondary reformer and an air compressor, where the primary reformer and/or the air compressor is operating at its maximum capacity, without investing to expand the capacity of either. A portion of the feed can bypass the primary reformer and enter the secondary reformer directly. Oxygen and nitrogen streams are added separately and independently to form an ammonia synthesis feed gas. Methane concentration in the outlet stream from the secondary reformer is controlled by manipulating the oxygen stream. Nitrogen concentration or a H2:N2 ratio in the ammonia converter feed gas is controlled by manipulating the nitrogen stream.

XYGEN AND NITROGEN INJECTION FOR INCREASING AMMONIA PRODUCTION

BACKGROUND OF THE INVENTION

1. FIELD OF THE INVENTION The present invention relates to an ammonia plant generating hydrogen using steam reformers. More particularly, the present invention relates to use of oxygen and nitrogen for increasing ammonia production.

2. DESCRIPTION OF THE RELATED ART An ammonia plant feeds hydrogen and nitrogen, and the hydrogen is generated by various methods including steam reforming using a primary reformer followed by a secondary reformer. A light hydrocarbon feed, typically methane, is fed to the primary reformer with steam where the feed reacts with steam over a catalyst in an endothermic reaction to form hydrogen, carbon monoxide, and carbon dioxide. Conversion is not complete, and the outlet gas from the primary reformer typically contains from 9 to 16 mole percent of unreacted methane. The outlet gas from the primary reformer is mixed with process air and fed to the secondary reformer where oxygen from the air burns with hydrogen and methane to provide heat for reforming reactions which take place over a fixed bed of catalyst. The outlet stream from the secondary reformer is an ammonia synthesis gas comprising hydrogen, nitrogen (introduced as a component of the process air), carbon monoxide, carbon dioxide, and some unreacted methane.

Carbon monoxide is not only a poison for an ammonia synthesis catalyst but it is also useless for the ammonia synthesis and therefore must be transformed. Thus, carbon monoxide is transformed into carbon dioxide and hydrogen through a two-stage shift reaction over a catalyst. The first stage is a high temperature shift reactor followed by a second stage, low temperature shift reactor. Carbon dioxide is thereafter removed from the produced gas by any adequate method. Remaining trace amounts of carbon monoxide and carbon dioxide are transformed in a methanation reactor to methane. Water is typically taken out by condensation. This provides a synthesis gas suitable for feed to an ammonia synthesis section. Often, an ammonia production plant is limited in production by the primary steam reformer and compressor capacity for providing the process air. Thus, the primary steam reformer and the air compressor are frequently bottlenecks in the ammonia production plant.

To produce more hydrogen for feed to an ammonia synthesis section, the primary steam reformer would need to accept a higher thermal load in order to provide the heat necessary for the endothermic reforming reaction. This is not possible where for various reasons the primary steam reformer is running at maximum capacity. Where the primary steam reformer can produce more hydrogen, then more air must be provided by the process air compressors.

However, this is not possible if the compressors are running at maximum capacity.

U.S. Patent No. 4,414,191 to Fuderer describes a process for the production of ammonia using a pressure swing absorption system for the purification of hydrogen to be used in an ammonia synthesis gas, where nitrogen is employed as a purge gas at an elevated purge pressure. The hydrogen recovered at absorption pressure contains about 20 to 25 per cent nitrogen, and is used as ammonia synthesis gas. An air separation system provides the nitrogen, and oxygen recovered from the air separation system is employed in a hydrogen generation system. Although the nitrogen is used beneficially for purging the pressure swing absorption beds, it is not controllably added for the primary purpose of providing nitrogen in a proper stoichiometric ratio with hydrogen for producing an ammonia synthesis gas.

U.S. Patent No. 4,681,745 to Pinto describes production of ammonia by a sequence of steam hydrocarbon primary reforming, secondary reforming with air, carbon monoxide shift conversion, carbon oxides removal and catalytic ammonia synthesis using oxygen-enriched air at the secondary reformer. The oxygen enriched air can be the by-product of a simple air separation plant producing nitrogen, where the nitrogen is used to aid start-up or shut-down of the process or to keep the process plant in a hot stand-by condition. An air separation plant provides a substantially pure nitrogen stream and a stream enriched in oxygen, with the nitrogen preferably stored and used as a purge. However, nitrogen is not controllably added for the purpose of providing nitrogen reactant in a beneficial ratio to hydrogen for increasing ammonia production.

U.S. Patent No. 5,180,570 to Lee et al. describes an integrated process for making methanol and ammonia from a hydrocarbon feed stock using an air separation unit to produce substantially pure oxygen and nitrogen gas streams. The oxygen gases used in the secondary reformer increase the operating pressure of the reformers so that compression to methanol synthesis pressure may be done in a single stage compressor. The nitrogen gas is used to remove carbon oxides impurities from an ammonia synthesis feed stream in a nitrogen wash unit in addition to supplying the nitrogen reactant in the ammonia synthesis gas. This patent is directed to an integrated process for making both methanol and ammonia rather than to increasing ammonia production by controlled addition of oxygen and nitrogen.

U.S. Patent No. 4,988,491 to Van Dijk et al. describes an integrated production of ammonia and urea using adiabatic reforming with substantially pure oxygen and nitrogen so that inert gases such as argon are not introduced into the system. With adiabatic reforming there is no primary reformer, and oxygen and nitrogen are not controllably added for debottlenecking an ammonia plant constrained by its primary reformer.

E.P. 770 578 published after the priority date to which the present invention is entitled, relates to ammonia production with enriched air reforming and nitrogen injection into the synthesis loop. Thus, there remains a need for systems and methods for debottlenecking an existing ammonia plant which is limited in production by its primary steam reformer or its process air compressors. When an ammonia production plant is limited by the primary steam reformer and/or the process air compressors, it would be highly desirable to increase the production of hydrogen for feed to an ammonia synthesis section without major expenditure to modify or replace this equipment. In such a desired process, production capacity should be increased while respecting the thermal constraints in the reformers, minimizing investment and optimizing the operation of the plant.

SUMMARY OF THE INVENTION

In an ammonia synthesis plant using primary and secondary steam reformers for hydrogen production, an air compressor supplies air to the secondary reformer, providing oxygen for combustion and heat generation in the secondary reformer and nitrogen for use in ammonia synthesis. The air compressor and/or the primary reformer is operating at maximum capacity. The present invention provides a process for increasing ammonia production without expanding the capacity of either the primary reformer or the air compressor, and thus without the need for such investment. Feed preferably (but not necessarily) bypasses the primary reformer and enters the secondary reformer directly. Preferably up to about 30 percent of total feed to the ammonia plant can be fed directly to the secondary reformer in this manner.

In accordance with the invention, an oxygen stream is added to enrich process air fed to the secondary reformer for forming a synthesis gas, which contains typically less than 0.6 percent methane molar on dry basis. Methane concentration or leakage in the synthesis gas is controlled by manipulating the flow rate of the oxygen stream. The secondary reformer is consequently more independent of the primary reformer than in traditional schemes, allowing smoother operation. Further, in accordance with the invention, a nitrogen-enriched stream is added to the synthesis gas to form a feed gas for an ammonia synthesis section. The ammonia synthesis feed gas produced this way contains typically less than 1.3% inerts (on dry basis).

The nitrogen concentration, or preferably a ratio of H2:N2, is controlled by manipulating the

flow rate of the added nitrogen stream. With independent control of the H2:N2 ratio, the

ammonia synthesis section is more decoupled from the steam reformers than in traditional schemes, which allows smoother operation. With the system and process of the present invention, ammoma production is substantially increased by separate and independent oxygen and nitrogen addition without investment in the primary reformer or the air compressor.

BRIEF DESCRIPTION OF THE DRAWING

A better understanding of the present invention can be obtained when the following detailed description is considered together with the figure, which is a schematic illustration of an ammoma synthesis plant with oxygen and nitrogen addition according to the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

To increase production of ammonia from an existing ammonia plant, which uses primary and secondary steam reformers, both hydrogen and nitrogen flows to an ammonia synthesis section must be increased. To increase the hydrogen flow, the primary reformer must accept higher thermal loads in order to provide the heat necessary for endothermic reforming reactions. This is not always possible due to limitations on heat flux through tubes filled with catalyst in the primary reformer or other concerns. Higher heat flux creates higher thermal stress, and thus an increase in production typically requires a revamp of the tubes or an additional primary reformer. Both of these options are costly.

Where the primary reformer can accept an increase in its throughput, its outlet process gas must be further reacted in the secondary reformer. For the secondary reformer to accommodate an increase more air should be brought to the secondary reformer for providing heat from combustion for the endothermic reforming reactions and thus producing more hydrogen and decreasing methane leakage at the outlet of the secondary reformer. The additional air introduced also provides the nitrogen needed for the production increase.

However, it may not be possible to bring more air into the secondary reformer because frequently the air compressor used to provide air is operating at its maximum capacity.

Typically, process air flow to the secondary reformer is manipulated to obtain the required H₂:N₂ ratio in the plant, and primary reformer firing (fuel gas inlet flow to the primary reformer) is manipulated to obtain a specified methane leakage (concentration) at the outlet of the secondary reformer. As primary reformer firing increases, methane leakage decreases. Methane leakage should be controlled because excessive methane leakage contributes to greater losses of reactants in a synthesis purge, thus smaller efficiency.

Sometimes a fair decrease of the methane leakage leads to a better process efficiency.

Nevertheless, in case that neither oxygen nor nitrogen are injected, this results, inevitably, to a higher heat load in the primary reformer tubes which leads to shorter life of tubes and catalyst. On the other hand, control of the H₂:N₂ ratio is essential to control the ammonia synthesis section, mainly the operation of the ammonia converter (for example, in cases where catalysts other than magnetite ones are used).

As mentioned before, in traditional schemes (without oxygen nor nitrogen injection) the process air flow rate to the secondary reformer is manipulated to obtain a desired H2.N2 ratio upstream of the ammoma converter. Nevertheless, it is frequent in industrial practice for operators to also manipulate the air flow rate to adjust the methane leakage because the oxygen contained in the air also influences this leakage. But this interferes with the H :N ratio. In other words, as the nitrogen to oxygen ratio in the air is fixed (N₂:O ratio = 3.71 mol/mol), bringing incremental oxygen to the system by air means bringing inevitably 3.71 times more nitrogen: this does not necessarily correspond to the optimum H :N ratio for the ammonia synthesis section. On the contrary if the H₂:N₂ ratio is controlled this determines the air flow rate and, consequently, the oxygen flow rate leading to a secondary reformer operation and a methane leakage which is not necessarily the optimum. It is evident that such control schemes are rather rigid. It would be useful to have N :O ratios that are not constrained by the air composition for each particular plant and thus increase the independence and flexibility of the reformers and the ammonia synthesis section. The best way to achieve this is through an independent oxygen and nitrogen injection as provided with this invention.

For example, with a specific plant having a methane leakage of 0.4 percent at the secondary reformer outlet and without oxygen nor nitrogen injection, the specific energy consumption can be decreased by 0.35 percent by increasing adequately the primary reformer's firing to have a methane leakage of 0.28 percent. The counterpart is that this makes the heat duty of the primary reformer to increase, resulting in a 5 °C increase of the tubes wall temperature. This can be detrimental to the tubes life or even destroy the same (tubes could creep under stress. On the contrary, in the same example, with the injection of oxygen and nitrogen, as provided with this invention, a similar reduction in the methane leakage and the specific energy consumption can be obtained by increasing, in this case by 13 percent, the injected oxygen flow rate without increasing the heat load in the primary reformer.

Ammonia production may be increased through simultaneous oxygen and nitrogen injection while adjusting operational parameters thus avoiding a primary reformer revamp, additional construction, or purchase of additional air compressors. Production can be increased by as much as much as 40 percent, provided that downstream equipment such as high-temperature shift (HTS) and low-temperature shift (LTS) reactors, carbon dioxide removal, the methanation unit, as well as some ammonia synthesis section units, have sufficient capacity. Pure (nitrogen-free) oxygen is injected in the secondary reformer (in practice mixed with the inlet process air) to burn with hydrogen and some methane to provide some additional heat for the endothermic reforming reaction and thus increase hydrogen production and, at the same time, control methane leakage at its outlet without increasing the firing of the primary reformer. However, the ammonia synthesis feed gas produced this way

is typically poor in nitrogen (H2:N2 ratio typically about 3.15 to 3.3).

Thus, pure (oxygen-free) gaseous nitrogen is injected upstream of the ammonia synthesis section to control this ratio at an optimum level. Besides easy H?:N ratio control, nitrogen injection at this point also helps to diminish pressure drops in the front end (i.e. upstream of the ammonia synthesis section). Other injection points can be used such as the secondary reformer, HTS or LTS reactors, the methanation reactor, the steam condensation unit or, potentially, in the ammoma synthesis section. Injection points are chosen according to the design specifics presented in each plant, and consideration should be given to the pressure availability of gaseous nitrogen as well as its purity. The present invention also provides for part of a fresh hydrocarbon feed to be shifted directly to the secondary reformer in order to decrease energy consumption. Approximately 0 to 45 percent and more typically 5 to 30 percent of the fresh feed can be bypassed around the primary reformer and fed directly to the secondary reformer, but this percentage has to be optimized for each particular case.

Turning now to the Figure, an ammonia production plant 10 is schematically illustrated according to the present invention. The process described is simplified, and one skilled in the art of ammonia plants will appreciate that although an ammonia plant is much more complex than described, the import of the present invention can be understood. A natural gas feed 12 is preheated in a coil 14 in the associated heat recovery section of a primary reformer 16, and sulfur compounds are removed in a desulfurization unit 18. A stream containing hydrogen (most often deviated from the ammonia synthesis section) is usually added upstream of this unit. Fumes 134 from a burner exchange heat in 132 with process gas upstream the desulfurization unit 18 and burner's fuel flow rate is manipulated to maintain a constant inlet temperature in unit 18. Desulfurized feed flows through a line 20 and a preheat coil 22 into primary reformer 16. A source of saturated steam 24 is superheated in a coil 26 and expanded in a steam turbine 28 which discharges part of the steam through a line 30 into line 20, where it is mixed with desulfurized feed in line 20. The rest of the steam is usefully used in turbines to drive compressor shafts, or used in sections such as the C0₂ removal section (stream 107). Fumes 136 from a side burner allow control of the steam superheating temperature; thus protecting turbine 38, and providing constant steam quality. Heat for endothermic reactions that take place in primary reformer 16 is provided by burning fuel gas 32 with combustion air preheated in a coil 36 and provided through a line 38. Primary reformer 16 is operated at maximum capacity as determined by one of many possible constraints such as maximum tube metal temperature.

Feed 12 is typically primarily methane which reacts with steam 24 over catalyst inside tubes in primary reformer 16 to produce hydrogen and carbon oxides in an endothermic steam reforming reaction to produce a primary reformer product stream 40. Conversion is incomplete, and primary reformer product steam 40 contains typically nine to sixteen mole percent of unreacted methane on dry gas basis as well as unused steam from stream 24.

Primary reformer product stream 40 is fed to a secondary reformer 42, which is typically an adiabatic catalytic reactor. Drawing from the atmosphere, process air 44 is compressed in a three-stage compressor 46a, 46b and 46c and cooled with intermediate coolers 48a and 48b, which is referred to generally as air compressor 46. The compressor's

shaft is usefully driven by a steam turbine. The discharge from air compressor 46 flows via a line 50 through a preheat coil 52 into secondary reformer 42, where process air 44 mixes with primary reformer product stream 40. Some hydrogen and some unreacted methane in primary reformer product stream 40 are combusted with the oxygen in process air 44 to provide heat for production of additional hydrogen in secondary reformer 42.

Air compressor 46 is operated at its maximum capacity as determined by any one of several possible constraints such as a metallurgically determined speed constraint. Secondary reformer 42 and downstream units have additional capacity so the limit on ammonia production is determined by primary reformer 16 and air compressor 46. In this scenario primary reformer 16 cannot provide additional primary reformer product stream 40, and air compressor 46 cannot deliver additional air 44 to secondary reformer 42. Primary reformer 16 and air compressor 46 are thus a bottleneck to further ammonia production. This problem is alleviated by the present invention. Air 54 is separated in an air separation unit 56 to produce a substantially-pure

(essentially nitrogen-free) oxygen stream 58 and a substantially-pure (essentially oxygen-free) nitrogen stream 60, or nitrogen or oxygen may be brought externally e.g., by pipeline, or obtained by any production means. Preferably, oxygen stream 58 has typically an oxygen concentration of greater than 99.5 mole percent, and nitrogen stream 60 has typically a nitrogen concentration of almost 100 mole percent (containing from 2-5 ppm atomic oxygen with all oxygenated compounds in it), however, the purity may be less. Oxygen stream 58 is added to process air line 50 for providing more oxygen in secondary reformer 42. With additional oxygen available, more hydrogen and methane can be burned in secondary reformer 42 to provide heat for a net increase in production of hydrogen in an ammonia synthesis gas 62. Since oxygen is no longer a limit to hydrogen production, desulfurized feed in line 20 can be fed directly to secondary reformer 42 via a bypass line 64. Water for steam reforming reactions in secondary reformer 42 is provided by, if necessary, increasing the flow rate of steam 24 so that sufficient steam is in primary reformer product stream 40 for reactions in secondary reformer 42. Another way of adding steam to the system, when increasing the capacity of the plant, is to increase flow rate 24, as above, but instead of adding this incremental flow to line 20 from line 30, some steam from line 30 can be directly added upstream or downstream of the secondary reformer bypassing the primary reformer. This strategy can help to avoid excessive pressure drops in the primary reformer while keeping the adequate steam to carbon ratio in the secondary reformer and the shift converters. Injecting some steam to the inlet of the secondary reformer will also allow to decrease secondary reformer's outlet temperature (the endothermic reforming reaction will progress more). Thus, by adding oxygen stream 58, the production of hydrogen in secondary reformer 42 can be increased, provided that inlet feed gas flow rate 12 is also increased.

Synthesis gas 62 comprises primarily hydrogen, nitrogen, carbon monoxide, carbon dioxide, and unreacted methane and steam. The concentration of unreacted methane, referred to as methane leakage, in synthesis gas 62 is controlled primarily by manipulating the flow rate of oxygen stream 58. This allows the desired methane concentration in the final ammonia synthesis feed gas 126 to be obtained. In an ammonia plant that does not have oxygen injection as described here, methane concentration in the synthesis gas 69 is controlled typically on the one hand by manipulating firing in the primary steam reformer, which is difficult because there is a great deal of lag time between manipulating fuel gas flow to the primary reformer and detecting an ultimate change in methane concentration in the synthesis gas 62.

With the present invention, primary reformer 16 is controlled to maximize the production of hydrogen in primary reformer product stream 40 by operating up against a constraint, whatever that constraint may be. In this mode primary reformer 16 is operated under relatively constant conditions, which allows for better utilization of the equipment, allowing operation at optimum conditions. Under these conditions, the variations in the quality of primary reformer product stream 40 are minimized. The feed, primary reformer product stream 40, to secondary reformer 42 is more stable since primary reformer 16 is now more decoupled from secondary reformer 42 because methane concentration in synthesis gas 62 is not controlled primarily by manipulating the flow rate of fuel gas 32 to primary reformer 16. Instead, methane concentration in synthesis gas 62 is controlled primarily by manipulating the flow rate of oxygen stream 58 while the firing of the primary reformer can be kept practically constant. By increasing the flow rate of the oxygen stream 58 the methane concentration in synthesis gas 62 decreases and by decreasing the flow rate of stream 58 the methane concentration (leakage) in synthesis gas 62 increases. This clearly allows to reach the methane leakage set-point in synthesis gas 62. This is a much simpler control strategy than has been used in the past, partly because the lag time between a change in flow rate of oxygen stream 58 and an ultimate change in methane concentration in synthesis gas 62 is relatively short as compared to a change in flow rate of fuel gas 32 and an ultimate change in methane concentration in synthesis gas 62.

Hydrogen production from secondary reformer 42 is increased until a constraint is reached, thus providing sufficient hydrogen production to accommodate the ammonia production increase in the following manner. Air compressor 46 is operated at maximum capacity to provide the maximum amount of air possible to secondary reformer 42 (air intake

flow rate 44 can not be further increased), while providing a relatively constant flow. Primary reformer 16 is operated at maximum capacity and thus partially reformed feed gas 40 to the secondary reformer 42 can not be further increased. Feed to secondary reformer 42 through bypass line 64 is increased until a limit is reached, which is typically a limit on secondary reformer 42 or on the downstream pieces of equipment. By bypassing feed around primary reformer 16 and feeding directly to secondary reformer 42 through bypass line 64, flow of feed 12 can be increased up to 45 percent and more typically by 5 to 30 percent. In another embodiment of the present invention the increase of feed gas 12 is not bypassed around primary reformer through line 64 but continues through line 20 and flows unreformed through primary reformer tubes, resulting in an increase of the methane concentration in line 40. Oxygen is brought into the secondary reformer 42 using line 58 and then 50 and its flow rate is increased until there is sufficient heat from combustion in the secondary reformer to reform the increase of feed gas 12, thus increasing hydrogen production to accommodate the ammoma production increase and decreasing the methane leakage in synthesis gas 62 to the appropriate level.

The synthesis of ammonia (NH3) requires one mole of nitrogen for three moles of

hydrogen. With the addition of substantially-pure oxygen stream 58 to process air line 50, the concentration of nitrogen in ammonia synthesis gas is less than it would be if only process air 44 was present in line 50, since nitrogen is a major component of air. Consequently, there is less than one mole of nitrogen per three moles of hydrogen (more than three moles of hydrogen per mole of nitrogen) also in ammoma synthesis gas 124. A ratio of hydrogen to

nitrogen (H2:N2) is monitored and expressed as moles of hydrogen per mole of nitrogen.

With the addition of oxygen stream 58, the H2:N2 ratio in ammoma synthesis gas 124

typically ranges between about 3.15 and about 3.30 before separate addition of nitrogen.

The concentration of nitrogen in ammonia synthesis gas 124 is increased, to form the

final ammonia synthesis feed gas 126, by adding substantially-pure nitrogen stream 60 to synthesis gas 62 to provide a desired ratio of hydrogen to nitrogen at the ammonia converter inlet. This arrangement is easy to control. Primary reformer 16 and secondary reformer 42 are operated to increase production of hydrogen to accommodate the ammonia production increase. The flow rate of nitrogen stream 60 is then simply manipulated to provide a desired concentration of nitrogen in the ammonia converter inlet gas. An online hydrogen analyzer and a flow meter is used to determine the amount of hydrogen upstream of the ammonia converter, and a simple ratio controller is used to manipulate the flow rate of nitrogen stream 60 to maintain a desired ratio of hydrogen to nitrogen.

Nitrogen stream 60 is illustrated here as being added to synthesis gas 62 just upstream of an ammonia synthesis section 66. This is the preferred injection point for nitrogen because it corresponds to the lowest pressure level of the ammonia synthesis gas and thus injected nitrogen can have lower pressure. On the other hand, injection of nitrogen at this point minimizes pressure drops in the synthesis gas preparation train (secondary reformer, shift reactors, methanation reactor) because nitrogen does not circulate in this section. This fact allows a higher pressure of the ammonia synthesis feed gas 126 resulting in smaller energy consumption in the compressors in the ammonia section. However, nitrogen stream 60 can be added at any point where it becomes a part of the feed to ammonia synthesis section 66. The point of addition is determined on a case-by-case basis, with practical consideration given to the available pressure and purity of the nitrogen, (e.g. nitrogen can be added upstream of the methanation reactor if it contains some oxygen, carbon oxides and water).

Downstream of secondary reformer 42, synthesis gas 62 is cooled in a boiler 70 for producing steam 72 and a product stream 74. Synthesis gas 62 contains carbon monoxide which is not useful for ammoma synthesis, and which is a poison for the ammonia converter

catalyst. Therefore, a high-temperature shift reactor 76 converts a portion of the carbon monoxide to carbon dioxide and hydrogen over a catalyst. High temperature shift reactor 76 produces a product stream 78, which is cooled in a boiler 80 producing steam 82 and a product stream 84 which is further cooled in an exchanger 86 heating a stream 88 (boiler feedwater) to produce a product stream 90. Still containing some carbon monoxide, product stream 90 is fed to a low temperature shift reactor 92 which shifts more carbon monoxide to carbon dioxide and hydrogen, producing a product stream 94.

Product stream 94 is further cooled in an exchanger 96 against a stream 98 (boiler feedwater) before being fed to a carbon dioxide recovery section 104, which removes carbon dioxide as a stream 106 using steam 107. A great part of the steam contained in the synthesis gas is also removed in this section. A product 108 is produced by carbon dioxide recovery unit 104 which is fed through a heat exchanger 110 into a methanation reactor 112 via a line

114. Further heating of stream 114 might be necessary by external means (e.g. in a heat exchanger using steam) to allow it to have the adequate temperature to render reaction over the methanation catalyst possible. Methanation product 116 is cooled in exchanger 118 to condense remaining steam which is removed in unit 122. The process gas from secondary reformer 42 through product stream 124 is referred to generally as synthesis gas. Nitrogen stream 60 is added to the process at some point to provide an ammonia synthesis feed stream 126 which is fed to ammonia synthesis section 66. A purge 128 is taken from ammonia synthesis section 66 and used as fuel (e.g. in the primary reformer burners) and an ammonia product stream 130 is produced.

In this manner ammonia production can be increased where the primary reformer and the air compressor would otherwise limit an increase in ammonia production. A need may remain for debottlenecking other portions of the plant, such as the high or low temperature shift reactors or the methanation reactor or units in the synthesis section, but in any case the investments spared by the present invention will represent a significant part of the total investments. The present invention is further illustrated by the following examples, which were developed using a computer simulation of an ammonia plant such as the one illustrated in the Figure and described above. A common basis is provided for each example, namely: 1. Ammonia production is increased from an initial 977.6 to 1075.4 tons/day (t/d), an increase of 97.8 t/d, a ten percent increase in production rate;

2. A side burner is used to control the temperature of the superheated steam fed to turbine 28. The fuel burned in the burner is included in the "total fuel gas" mentioned in the Table.

3. The fuel to an auxiliary boiler is manipulated in order to close the steam and mechanical energy balance of the plant (steam by heat recovery plus steam by auxiliary boiler - process steam - steam used in turbines - steam export = 0). The electric power needed for an auxiliary air compressor (as defined in Example 1) is not included in this balance. The fuel to the auxiliary boiler is included in the "total fuel gas" mentioned in the Table;

4. Natural gas composition is (in mole percent): 87.74 % methane; 7.11 % ethane; 1.55 % carbon dioxide and 3.60 % nitrogen;

5. Oxygen stream composition is (in mole percent): 99.5 % oxygen and 0.5 % argon (although lower purity can be used in implementing the invention, but argon content should not be excessive);

6. Maximum secondary reformer outlet temperature is 1000°C for the sake of the

demonstration examples; and

7. Feed to an ammonia synthesis converter in the ammonia synthesis section contains 12 mole % inerts. This is an example: any other inerts ratio that respects constraints and optimizes the synthesis section operation can be used.

8. In the table, "GJ" is GJ Higher Heating Value and tons are metric tons. The power needed for oxygen and nitrogen production is not included in the "specific energy consumption", which is expressed in GJ/t NH3 in the Table.

The table below provides data for the following six examples, as well as for the initial case (prior to the capacity increase, called "Base"): TABLE

Examples

Base 3 4

Ammonia production 977.6 1075.4 1075.4 1075.4 1075.4 1075.4 1075.4 (t/day)

Operating cost parameters Units

Specific energy GJ/t NH3 37.256 37.741 37.566 37.264 37.945 37.103 36.737 consumption

Oxygen addition kg mol/hr 41 22 122 95 87

Nitrogen addition kg mol/hr 115 141 113 126 152

Auxiliary air compressor kW 673 - - - electricity

Investment in:

Primary reformer no yes yes yes no no no

Auxiliary air compressor no yes no no no no no

Process parameters:

H2:N2 ratio at ammonia 3.00 3.00 3.00 2.36 3.00 3.00 2.23 converter inlet

Feed bypass to sec. % mol 11.64 11.36 reformer

Total air How rate kg mol/hr 1578 1735 1578 1578 1578 1578 1578

(additional O₂ not included)

Heat flux in prim. Reformer kW/m² 62.3 68.9 68.9 68.7 62.3 62.3 62.3 tubes

Ammoma synthesis feed % mol (dry) 1.06 1.07 0.97 1.07 0.97 1.12 1.15 gas inerts

Ammonia Synthesis feed bar 27.48 25.79 26.07 26.16 25.46 27.00 27.02 gas pressure

Total mechanical power kW 29149 32839 32466 31820 32960 32017 31306

Steam to prim. Reformer ton/hr 71.8 79.0 78.9 78.6 81.5 71.8 71.8

Total fuel gas GJ/t NH3 12.886 13.364 13.214 12.989 12.801 12.020 11.731

Process gas (Feed) GJ/t NH3 24.370 24.377 24.352 24.275 25.144 25.083 25.006

Air enrichment (Oxygen % mol 21.00 21.00 22.97 22.08 26.62 25.45 25.09 concentration)

Methane leakage primary % mol (dry) 1 1.1 1 11.13 1 1.1 1 11.11 15.05 11.12 11.12 reformer

Methane leakage secondary % mol (dry) 0.30 0.30 0.24 0.34 0.25 0.29 0.32 reformer

Outlet temp, secondary ref. °C 978 975 990 971 981 993 986

Synthesis compressor speed rpm 10028 10649 10577 10448 10695 10436 10293

CO₂ export kg mol/hr ^{1 1 18} 1230 1233 1223 1275 1262 1257

EXAMPLE 1

This is a case for increasing ammonia production from the initial 977.6 to 1075.4 t/d without injecting oxygen or nitrogen. Production would be increased by either revamping the primary reformer, such as by installing thinner tubes or more active catalyst, or by providing an additional primary reformer in parallel to the existing one. An existing air compressor works at its upper limit (1,578 kg mole/hr) and is driven by a steam turbine. An auxiliary air compressor would be purchased and placed in parallel with the existing one. The total inlet

air flow rate is adjusted to satisfy nitrogen requirements, maintaining a H2:N2 ratio of 3.00, as

can be seen in the Table. An electric motor would be used to drive the auxiliary air compressor because it is more flexible than a steam turbine, and electric is preferred for small air flow rates. With reference to the Table, its energy consumption is 673 kW and is not in the steam-mechanical energy balance.

EXAMPLE 2 Ammonia production is increased to 1075.4 t/d as is the case for all of the examples.

While in Example 1, production was increased by both revamping the primary reformer (or adding a new one) and adding an auxiliary air compressor, that is not the case here. In Example 2 the primary reformer is again assumed revamped, but the auxiliary air compressor is eliminated. The heat flux in the primary reformer tubes is the same as in Example 1 at 68.9

2 KW/m . A single air compressor is working at its upper limit, providing a flow rate of 1,578

kg mol/hr as compared to 1,735 kg mol/hr in Example 1 with its auxiliary air compressor. Oxygen was added at the rate of 41 kg mol/hr to process air line 50 via oxygen stream 58. A

H2:N2 ratio of 3.00 was maintained between Examples 1 and 2 by injecting nitrogen as shown

in the Figure. The flow rates of oxygen stream 58, nitrogen stream 60, feed 12 and fuel gas 32 were adjusted to accommodate the production increase within operating constraints while minimizing operating cost.

Since there is no auxiliary air compressor, the air flow rate is smaller in Example 2 than in Example 1. Pressure drops are thus smaller, and ammonia synthesis feed gas 126 pressure is higher (26.07 bar instead of 25.79 bar). As compared to Example 1 the total mechanical power requirement is reduced from 32839 to 32466 kW and the total fuel gas

requirement is reduced from 13.364 to 13.214 GJ/t NH3. Thus, specific energy consumption

decreases from 37.741 to 37.566 GJ/t NH3. Not only is the investment for the auxiliary air

compressor eliminated, the operating cost of 673 kW is also eliminated. This is illustrative of a situation where an ammonia plant cannot increase production because its air compressor is operating at maximum capacity while other portions of the plant can increase production.

Example 2 thus shows that if production can otherwise be increased, then investment in and operating costs for an auxiliary air compressor can be eliminated by adding oxygen and nitrogen instead according to the present invention.

EXAMPLE 3

In this simulation the operating configuration is the same as in Example 2 (production

bottleneck is the air compressor), but the H2:N2 ratio upstream of ammonia converter in

section 66 is optimized to further reduce operating cost. The air flow rate here is the same as in Example 2, which is less than in Example 1 since no auxiliary air compressor is installed for the ammonia production increase. However, the flow rate of oxygen stream 58 has been decreased from 41 to 22 kg mol/hr, allowing methane concentration in ammonia synthesis gas 62 to increase from 0.24 to 0.34 mol %, and the flow rate of nitrogen stream 60 has been increased from 115 to 141 kg mol/hr. Example 3 provides a more cost effective solution than Example 2 since nitrogen is less expensive than oxygen and the specific consumption decreases as explained in the paragraph below.

Pressure drops downstream of secondary reformer 42 are reduced resulting in a higher pressure for ammoma synthesis feed gas 126. The pressure is 26.16 instead of 26.07 bar, as

compared to 25.79 bar in Example 1. The H2:N2 ratio is reduced from 3.00 to 2.36. For the

specific case the ammonia converter performs better and less matter circulates in the ammonia synthesis section 66 to produce the same amount of ammoma. As compared to Example 2 the total mechanical power is reduced from 32466 to 31820 kW and total fuel gas consumption is down to 12.989 from 13.214 GJ/t of ammonia produced. Overall, specific energy consumption is reduced from 37.566 to 37.264 GJ/t of ammonia produced.

This case demonstrates the benefits of dissociating the oxygen and nitrogen injection in an ammoma plant. The flow rate of nitrogen stream 60 is easy to manipulate, and the

H2:N2 ratio can be easily controlled without perturbing the operation of the primary and

secondary reformers, which allows the optimization of the H2: 2 ratio in ammonia synthesis

section 66. On the other hand oxygen injection allows independent control of the methane

leakage (concentration) downstream of secondary reformer 42 in ammonia synthesis gas 62. In the specific case examined, it is more economical to have a greater methane leakage (0.34% instead of 0.24 mol % for Example 2 or 0.30 mol % for Example 1). This demonstrates the flexibility that separate oxygen and nitrogen injection provides in an ammonia plant. This is a preferred case, when the air compressor is the bottleneck, the primary reformer either having sufficient capacity or an investment being made in it to provide sufficient capacity for a ten percent increase in ammonia production. EXAMPLE 4

While Example 3 was used to illustrate an incremental improvement over Example 2,

Example 4 should be compared to Example 1. In Example 4 there is no investment in either the primary reformer or the air compressor, yet ammonia production is increased by ten percent, which is the basis for all of the examples. Both the primary reformer and the air compressor are assumed to be operating at maximum capacity. The heat flux in the primary

2 reformer was not allowed to exceed 62.3 kW/m , which corresponds to the heat flux in the

primary reformer for the initial plant capacity of 977.6 t NH₃/day (base example). There is no

auxiliary air compressor, and air compressor 46 provides 1578 kg mol/hr of process air while consuming 6737 kW of energy. The flow rate of oxygen stream 58 is 122 kg mol/hr, and the flow rate of nitrogen stream 60 is 113 kg mol/hr versus none of either in Example 1. The

H2:N2 ratio is 3.00 which is the same as in Example 1.

The flow rates of oxygen stream 58, nitrogen stream 60, feed 12, and fuel gas 32 are adjusted to provide the ammonia production increase while satisfying various constraints, such as maximum heat flux in primary reformer 16, and minimizing operating costs. There is no flow in bypass line 64, so all of the process gas feed to secondary reformer 42 is provided by primary reformer product stream 40.

In this case, due to the limitation in the heat flux in the primary reformer tubes, the gas passing through it is less reformed, i.e. the methane leakage in the primary reformer is 15.05 versus 11.13 mol % for Example 1. Thus, the process gas has to be further reformed in the secondary reformer. In comparison with Example 1 , additional process gas has to be fed to the secondary reformer because part of it is used as fuel gas to provide heat by combustion. Hence, the pressure drops increase because this gas is passing through the primary reformer where the pressure drops are more important than in the secondary reformer. Due to a constraint on the steam-to-carbon ratio (3.5:1), more steam has to be mixed with the process gas and this contributes to the pressure drop increase as well as requiring greater steam production. On the other hand, less fuel gas per ton of ammonia produced is needed for firing the primary reformer because the heat flux in the primary reformer remains as in the base example while production increases.

Globally, the decrease in fuel consumption (12.801 GJ/t instead of 13.364 GJ/t) is less important than the increase of the process gas consumption (25.144 GJ/t instead of 24.377 GJ/t NH₃ produced) and therefore the specific energy consumption increases to 37.945 from 37.741 GJ/t of ammonia produced. Because more gas is now introduced in the secondary reformer to be reformed, the amount of pure oxygen added is significant (122 kg mol/hr,

enriching the air to 26.62 mol % O₂). This increases the operating cost for this case, although

any investment to revamp the primary reformer and for an auxiliary air compressor is avoided. Although ammonia production is increased, the primary reformer operates at conditions even smoother than when the plant produced 977.6 t/day of ammonia because the leakage is higher than in the base example. Thus, production can be increased without placing a thermal stress on the primary reformer so the tubes and the catalyst will last longer.

For this case the CO export increases (1275 kg mol/hr instead of 1230 for Example 1). This is a supplementary benefit.

EXAMPLE 5

In this case feed bypasses the primary reformer putting 11.64 mol % of feed 12 into bypass line 64 for direct feed to secondary reformer 42. This amount is chosen so as to minimize operating cost without increasing excessively the secondary reformer's outlet temperature, staying below the 1000 °C constraint but increasing from 975 to 993 °C. The

flow rate of oxygen stream 58 is 95 kg mol/hr, and the flow rate of nitrogen stream 60 is 126 kg mol/hr versus none of either in Example 1. The H2:N2 ratio is 3.00, the same as in

Example 1.

Some of the process gas is used now in the secondary reformer as fuel gas to provide

heat by combustion. Consequently, more process gas is used (25.083 versus 24.377 GJ/t NH₃

for Example 1). Steam for reforming reactions in secondary reformer 42 is provided by steam passing through the primary reformer, and its quantity is sufficient to guarantee a trouble-free operation in the secondary reformer. In fact the amount of steam required is reduced from 79.0 ton/hr steam for the primary reformer in Example 1 to 71.8 t/hr in this case. This makes the fuel gas consumption to decrease. If the steam were not sufficient, steam can be added to the process gas going directly to the secondary reformer, or just upstream the high- temperature shift reactor. Because smaller amounts of process gas and steam are passing through the primary reformer due to the bypass, and less nitrogen is present in the front end, due to the nitrogen injection just before the ammonia synthesis section, the pressure drops are significantly lower. Consequently, the inlet pressure to the synthesis section is higher, 27.00 bar instead of 25.79 for Example 1, and power consumption in the plant is now 32017 versus 32839 kW for Example 1. Also, less fuel gas per ton of ammonia produced is needed for firing the primary reformer because the heat flux in the primary reformer remains as in the base example while production increases. For all the reasons mentioned above, fuel gas is

reduced to 12.020 from 13.364 GJ/t NH₃ produced. Total specific energy consumption is

reduced from 37.741 to 37.103 GJ/t NH₃ produced.

Thus, although ammonia production is increased 10 % without an investment in either the primary reformer or the air compressors, total specific energy consumption is lower than if these investments were made. This illustrates that for an ammonia plant constrained by its primary reformer and its process air compressor, addition of oxygen and nitrogen allows ammonia production to be increased without requiring an investment in either the primary reformer or the air compressor.

Although the capacity is increased by 10%, the primary reformer suffers the same thermal stress (62.3 kW/m2 as in the base example (operation of 977.6 t/day). Also, the CO₂ export increases (1262 kg mol/hr instead of 1230 kg mol/hr) and this is a supplementary benefit.

EXAMPLE 6

In this case the operation illustrated in Example 5 is further optimized. Again, both the primary reformer and the air compressor are assumed to be operating at maximum capacity. There is no investment in either the primary reformer or the air compressor, and ammonia production is increased by ten percent. Less oxygen is used here than in Example 5,

and more nitrogen is used. This reduces the H2:N2 ratio which reduces operating costs below

that in Example 5. Operating costs are lower for the following reasons. As compared to Example 5, oxygen consumption is lower at 87 rather than 95 kg mol/hr. In the specific case the ammonia converter works more efficiently for lower H2:N2 ratio and, thus, the flow rate in the synthesis section is lower while producing the same amount of ammonia. The power consumption is lower (31306 versus 32017 in Example 5 versus 32839 kW in Example 1), and the fuel gas consumption is lower (11.731 versus 12.020 in Example 5 versus 13.364 GJ/t

NH₃ in Example 1). Thus the specific energy consumption decreases (36.737 versus 37.103 in

Example 5 versus 37.741 GJ/t NH₃ in Example 1), and this provides lower operating costs.

Again, as in Example 5, 673 kW of electrical power are not needed since the auxiliary compressor is not needed. Although the capacity is increased by 10% the primary reformer suffers the same thermal stress (62.3 kW/m2) as in the base example (operation of 977.6 t/day).

The CO₂ export also increases (1257 kg mol/hr instead of 1230 kg mol/hr) and this is a supplementary benefit. This is a preferred case when both the primary reformer and the air compressor are bottlenecks to an increase in ammonia production. A portion of the feed (11.36 mol % of feed

12 ) bypasses the primary reformer, and is fed directly to the secondary reformer via bypass

line 64. By easily manipulating nitrogen flow, the H2:N2 ratio is optimized for the plant.

Oxygen injection is easily manipulated not only to produce more hydrogen but also to control methane leakage without excessively increasing the secondary reformer's outlet temperature, and its operating cost is reasonable in light of the increase in ammonia production without investment in the primary reformer or the air compressor.

With oxygen and nitrogen injection, the rotation speed of an ammonia synthesis gas compressor is reduced (10293 versus 10649 rpm in Example 1). This means that the invention proposed can help to overcome another frequently encountered bottleneck, the ammonia synthesis gas compressor.

It is therefore possible to increase the capacity of an ammonia production plant fed by steam reforming where the primary steam reformer and the process air compressor are bottlenecks without investing in supplementary primary steam reformer capacity or in an auxiliary air compressor. The invention does not change substantially the operating conditions of the primary reformer, which allows for a longer life for this piece of equipment. Oxygen, separately injected, provides the necessary heat for reforming in the secondary reformer by combustion with process gas there, and methane leakage can be easily controlled by manipulating the amount of oxygen injected. Nitrogen injection supplements the supply of this necessary reactant for an increase in ammonia production. The H2:N2 ratio is easily

optimized by simply manipulating the flow of nitrogen stream 60. Separate and independent nitrogen injection decouples the steam reformers from the ammoma synthesis section to a great extent, thus smoothing out the operation of the plant. Similarly, control of methane leakage by manipulating the flow of oxygen stream 58 decouples the primary reformer from the secondary reformer to a greater extent than in traditional schemes, thus providing a smoother operation. Thus, control schemes can be more easily and successfully implemented to optimize hydrogen production from each reformer, controlling and manipulating multiple variables to maintain each reformer at its operating constraint. The present invention thus allows ammonia production capacity to be increased without significant capital investment while decreasing the specific energy consumption of the plant and increasing its operating flexibility.

The foregoing disclosure and description of the invention are illustrative and explanatory thereof, and various modifications and alterations to the embodiments disclosed herein will be apparent to those skilled in the art in view of this disclosure. It is intended that all such variations and modifications fall within the spirit and scope of this invention as claimed.

Claims (English, EP 0914294 B1) Jump to:   

1. A process for increasing ammonia production in an ammonia synthesis plant having a primary reformer, a secondary reformer and an air compressor for supplying air to the said secondary reformer, without expanding the capacity of either the primary reformer or the air compressor, the process comprising :
• flowing a light hydrocarbon feed to the ammonia synthesis plant,
• flowing a secondary reformer feed to the secondary reformer,
• adding an oxygen stream to the air feeding of the secondary reformer for forming synthesis gas containing methane, the oxygen stream having a purity of at least 90 mole percent oxygen,
• controlling the concentration of methane in the synthesis gas by manipulating the flow rate of the oxygen stream,
• adding a nitrogen stream to the synthesis gas to form an ammonia synthesis feed gas, and the nitrogen stream having a purity of at least 90 mole percent nitrogen, and
• controlling a H₂:N₂ ratio at an ammonia converter inlet by manipulating the flow rate of the nitrogen stream.
2. The process of claim 1, wherein at least a portion of the light hydrocarbon fend is fed directly to the secondary reformer as at least a portion of the secondary reformer feed, said portion bypassing the primary reformer.
3. The process of claim 1, wherein the increase in ammonia production requires an increase in the light hydrocarbon feed, and a major portion of the increase in this lighthydrocarbon feed is fed directly to the secondary reformer as at least a portion of the secondary reformer feed.
4. The process of claim 1, wherein the purity of the oxygen stream is at least 97 mole percent.
5. The process of claim 1, wherein the purity of the nitrogen stream is at least 97 mole percent.
6. The process of claim 1, wherein the H₂:N₂ ratio at the ammonia converter inlet is controlled to less than or equal to 3.0.
7. The process of claim 1, wherein the H₂:N₂ ratio at the ammonia converter inlet is greater than 3.0.