In steam reforming, hydrogen is produced by reforming the hydrocarbon feedstock, producing synthesis gas containing a mixture of carbon monoxide and hydrogen. The carbon monoxide is then reacted with steam in the water-gas-shift reaction to produce carbon dioxide and hydrogen. The carbon dioxide is recovered for urea production, exported as co-product, or vented to the atmosphere. In the final synthesis loop, the hydrogen reacts with nitrogen to form ammonia.
In steam reforming ammonia plants there is a surplus of high-level heat that is produced in primary reforming, secondary reforming, shift conversion and ammonia synthesis. Most of the waste heat is recovered for producing high pressure steam that is used in turbines for driving compressors, pumps and fans. In general, all the high pressure steam will be used in steam turbines for driving the synthesis gas compressor. Modern steam reforming ammonia plants do not import energy for driving the mechanical equipment. Energy is in many cases exported in the form of steam or electricity to other consumers (IPTS/EC, 2007).
The natural gas use in a an ammonia plant using the steam reforming process ranges between 28 and 35.5 GJ/tonne, of which about 20-22 GJ/tonne of ammonia is used as feedstock, 7.2-9.0 GJ/tonne is fuel consumed in the primary reformer, and the remaining 0.5-4.2 GJ/tonne is used in auxiliary boilers and others. Table below shows the typical breakdown of energy use in steam reforming ammonia plants.
| Natura Gas|
| Heat Input/Output|
|Primary reformer feed||20.4 - 22.3|
|Primary reformer||7.2 - 9.0||3.0 - 4.5|
|Waste heat boiler||-5.0 - -6.0|
|Shift and CO2 removal||0.8 - 1.2|
|Synthesis loop||-2.5 - -3.0|
|Auxiliary boiler||0.3 - 3.5||-0.2 - 3.0|
|Turbines/compressors||3.9 - 6.3|
|Other (e.g. flare||0.2 - 0.7|
|Total||28.1 - 35.5||0.0|
There are also a number of steam reforming configurations offered by different technology providers, and currently considered as best available technologies. These include the following:
- Advanced conventional primary reforming with high duty primary reforming and stoichiometric process air in the secondary reformer. Processes with this configuration are the Kellogg Low-Energy Ammonia Process, the Haldor Topsoe process, the Uhde process, the LEAD process, the Exxon Chemical Process, the Fluor process and the Lumus process (Ullmann’s, 2011).
- Steam reforming with mild condition in the primary reformer and use of excess air in the secondary reformer. Processes with this configuration are the Braun Purifier process, the ICI AMV process, the Foster Wheeler AM2 process, the Humphreys & Glascow BYAS process, the Jacob Plus Ammonia Technology the Montedison Low-pressure process and the Kellogg’s LEAP process (Ullmann’s, 2011).
- Heat exchange autothermal reforming with a process gas heat exchange reformer and a separate secondary reformer, or in combination with an autothermal reformer that uses excess or enriched air. Processes with this configuration are the ICI LCA process and the Chiyoda process (Ullmann’s, 2011).
There are also a number of processes in which there is no use of a secondary reformer in which the nitrogen is supplied by an air separation unit. Some examples are the Linde LAC process and the Humphreys & Glasgow MDF process. Claimed energy use values of the different processes are provided in Table below.
|Process Name|| Energy Use|
|Advanced conventional primary reforming|
|Kellogg Low-Energy Ammonia Process||27.9 (27.01)|
|Haldor Topsoe Process||27.9|
|Uhde Process||28.0 (27.0)|
|Exxon Chemical Process||29.0|
|Processes with reduced primary refiner firing|
|Braun Purifier Process||28.0 (27.01)|
|ICI AMV Process||28.5|
|Foster Wheeler AM2 Process||29.3|
|Humphreys & Glasgow BYAS Process||28.7|
|Jacobs Plus Ammonia Technology||28.8 (26.81)|
|Montedison Low-Pressure Process||28.1|
|Kellogg’s LEAP Process||<28.0|
|Processes without a primary reformer|
|ICI LCA Process||29.3|
|Kellogg Brown and Root (KBR) KAAPplus Process||27.2|
|Processes without a secondary reformer|
|KTI PARC Process||29.3-31.8|
|Linde LAC Process||28.5 (29.32)|
|Humphreys & Glasgow MDF Process||32.82|
1: Energy use when steam is exported
2: Energy use when CO2 is recovered.
Conventional steam reforming of natural gas includes desulphurization, primary reforming and secondary reforming processes.
The feedstock used for the production of ammonia may contain sulphur and sulphur compounds which are harmful for the catalyst used in subsequent process steps and therefore need to be removed. Typically, feedstocks may contain up to 5 mg S/Nm3 of sulphur compounds. In desulphurization, the pre-heated (350-400oC) and untreated feed-gas enters a vessel that usually contains a cobalt molybdenum catalyst where the sulphur compounds are hydrogenated to H2S. The hydrogen needed for the reaction is usually recycled from the synthesis section. The hydrogenated sulphur compounds are then adsorbed on pelletized zinc oxide. After desulphurization, the feed-gas sulphur concentration drops to less than 0.1 ppm.
The feed-gas, after being treated in a desulphurization vessel, is mixed with process steam. The preheated mixture enters the primary reformer at a temperature of 400-600oC. In certain new and revamped ammonia plants the preheated steam/gas mixture is passed through an adiabatic pre-reformer before entering the primary reformer, where it is then reheated in the convection section (EFMA, 2000).
Primary reformers consist of a large number of high-nickel chromium alloy tubes which are filled with a nickel-containing reforming catalyst (see Figure 3). In conventional steam reforming, the hydrocarbon conversion rate in the primary reformer is about 60%. The reaction is highly endothermic:
Natural gas or other types of burners provide heat to the process. About half of the heat supplied to the primary reformer is consumed in the reaction. The remaining half is contained in the flue gases and used in the convection section of the reformers for the preheating of process streams. The flue gases leaving the primary reformer convection section compose the most significant source of the plant’s emissions. These emissions mainly consist of CO2, NOx, and small amounts of SO2 and CO (EFMA, 2000).
The typical fuel use in the primary reformer (including steam generation) ranges between 7.2 and 9.0 GJ/tonne of ammonia (IPTS/EC, 2007). Natural gas consumption in energy efficient ammonia plants is about 6.8 GJ/tonne of ammonia (Ullmann’s, 2011).
In secondary reforming the nitrogen needed for the production of ammonia is added and the reforming of the hydrocarbon feed is completed (only about 60% of the feed-gas was reformed in the primary reformer). In order to increase the conversion rate, high temperatures are required. This is achieved with the internal combustion of part of the reaction gas and the process air, which is also the source of nitrogen, before it passes over to the nickel containing catalysts. The supplied air is compressed and heated at the convection section of the primary reformer at a temperature of 500-600oC. The gas outlet temperature is about 1000oC and about 99% of the primary reformer hydrocarbon feed is converted. The residual methane content is about 0.2-0.3% (dry gas base). Heat is removed with the use of a waste heat boiler and the gas is cooled down to approximately 330-380oC (IPTS/EC, 2007).
This paper investigates the technical and economical feasibility of several concepts for a 30 % capacity increase of an old ammonia plant. It shows an interesting way to overcome the limitations in the two most critical plant units: Reforming capacity is increased by a newly added autothermal reformer, while capacity is added to the ammonia synthesis by the Uhde Dual Pressure Process.
Using experience from reference projects, this process concept is compared to other technical options and is discussed on the basis of investment and operating cost.
One of the factors making this concept competitive is the fact that by installation of parallel equipment with few tie-ins only, the shutdown time for its implementation is very short. Another interesting feature is that the ATR concept offers more CO2 as a pure stream to be used in a urea plant compared to the other concepts. This provides the possibility to easily combine it with a larger urea plant.
A capacity increase of an ammonia plant can be a successful way to increase its economic viability. Also, a revamp involves considerably less risk than the erection of a new plant since the overall investment is moderate and project implementation takes less time.
A capacity enlargement up to about 10 to 15 % can usually be realized with moderate modifications by mobilizing the reserves which are already present in the majority of the process units. Only some equipment items are acting as bottlenecks and require modifications or re-placement.
Larger capacity increases tend to require more substantial measures and bigger changes in the process. As this makes the capacity increase considerably more expensive, it is of key importance to select the most cost effective solution.
Scope of the Study
Basis for the investigation is an existing ammo-nia plant in Russia. Its actual capacity at the time of preparing the study was 1680 mtpd. An expansion target of 30 % extra capacity was chosen. The plant had already undergone several capacity enlargements and modifications. Its current production capacity is considerably larger than its original nameplate capacity. Therefore, no significant potential for additional capacity is available in the existing plant equipment.
Three different revamp concepts are presented, with their main difference being within the reforming section. To compare their economic viability, capital and operating cost have been determined and evaluated for each revamp concept , .