M.ABDEL AATY, Misr Fertilizer Production Company, Damietta, Egypt
M.LUNN, Y. KAMAL, A. JONCKERS, Johnson Matthey Catalysts, Billingham, United Kingdom
This article was first published at Nytrogen and Syngas Conference, 21-24 February 2011, Düsseldorf, Germany
For over 100 years magnetite-based synthesis catalysts have formed the basis of ammonia synthesis. One of the most recent companies to join the rich heritage of the industry is Misr Fertilizer Production Company (MOPCO). MOPCO’s operating experience at one of the recent projects using the latest ammonia synthesis catalysts is described. Given the critical importance of reliable ammonia synthesis catalyst, investments have recently been made in ammonia synthesis testing and production technology, as this paper describes.
MISR FERTILIZER PRODUCTION COMPANY (MOPCO)
Misr Fertilizer Production Company (MOPCO) is an Egyptian fertiliser company owned mainly by the Egyptian Oil sector. MOPCO specializes in fertilizer production including urea and ammonia.
The MOPCO site is located inside the public free zone in Damietta on a 400,000 square meters piece of land zoned for the MOPCO project and its future expansions. Damietta is considered as one of the most important cities in Egypt and, nowadays, as one of the important industrial cities in the world due to the export of natural gas and petrochemical products through its port facilities which are conveniently located on the Mediterranean Sea. From this location it can readily service export markets with urea and has key customers located in North America (Canada, USA and Mexico), Latin America (Argentina), Europe (France, Greece, Turkey, Italy, Germany, England, Spain, Bulgaria and Ireland) and Africa (Angola and Mauritius) as well as Austrailia and India.
The MOPCO site has been in operation since 2008 and currently has a capacity of 1,200 mtpd ammonia (Uhde technology), 1925 mtpd of urea (Stamicarbon technology) and 2,000 mtpd granulation (Hydro Agri technology). Most of the ammonia is used to produce urea; the remaining amount (around 40,000 MTPY) is sold to the local market. The total investment for the first production line commissioned in 2008 was USD 400 million. The feedstock for the site is natural gas from the MOPCO offshore field, which offers a stable and long-term supply for at least 20 years. With the selected Uhde ammonia technology using catalysts supplied by Johnson Matthey Catalysts (UK), it is possible to achieve an expected consumption figure of 7.288 Gcal/t NH3.
Further expansion is planned with another two identical 1,200 mtpd Uhde ammonia plants and two more 1,925 mtpd Stamicarbon urea plants expected to start up in January and March, 2012. The two trains are currently under construction and the total investment cost for the expansion is USD 1.7 billion.
The pre-commissioning activities started 27 months after the effective date and began with the off–site facilities in November 2007 with the preparation of the instrument air and cooling water systems. After the package boiler was started up, steam blowing started on 22 April 2008 and was finished within two weeks. In the middle of May 2008 the reformer was lit for dry-out and the CO2 removal section was prepared by chemical cleaning. Low-temperature shift catalyst reduction was completed on 04 June 2008 and the ammonia converter catalyst was reduced in August 2008.
The commissioning of the ammonia plant started in May 2008 with the introduction of feed gas to the reformer to start the reduction of the front end of the unit containing the synthesis gas production catalysts. With the reduction of the synthesis catalyst completed, the first ammonia was sent to storage in August 2008 and the first urea was produced on 14 August 2008.
Before the plant was ramped up to full capacity, it was operated at reduced load (75% of design) for a time to fine tune the plant. During this period of initial operation MOPCO ensured that the plant could be operated continuously under commercial conditions once the test run was completed and that no further shut-down would be required until the next scheduled turnaround.
Performance test run
The performance test run was conducted in seven consecutive days during September 2008 on the whole integrated plant, including the related off-site facilities.
The plant performance test proved that the ammonia, urea synthesis and granulation plants, and all off-site facilities, could be operated simultaneously and continuously to produce final, intermediate products and by- products to specification at the guaranteed production rates. Furthermore, the consumption of raw materials, utilities, chemicals and other consumables were all within guaranteed figures.
Some extracts of the performance and consumption figures are given below.
The average production rate for the ammonia plant is 1,262 mtpd, which is 105.17% of nameplate capacity. The average NH3 concentration has been 99.8 min wt-%. The average efficiency is 7.243 Gcal/t NH3, which is better than the expected performance. In 2010 the unit has been on line for 346 days versus 330 days planned, which corresponds to a reliability of 104.8%.
OVERVIEW OF AMMONIA SYNTHESIS CATALYST
By the early 20th century it was recognised that demand for nitrogen fertilisers would far exceed that available from natural sources. In 1908, an intensive search began for a catalyst to produce synthetic ammonia from hydrogen and nitrogen. Studies by Haber and by Bosch and Mittasch at BASF identified promising iron, osmium, uranium and uranium carbide catalysts. Mittasch returned to iron-based materials and observed that small amounts of impurities greatly affected the performance of pure iron catalysts. In a comprehensive study of multicomponent catalysts, elements were identified that had either positive or negative effects on the performance of iron as a catalyst (see Table 1).
The critical discovery, made after a magnetite ore containing potash and alumina showed relatively high activity, was that a combination of an acidic or amphoteric oxide such as alumina, silica or zirconia with an alkaline oxide such as potash gave catalysts with high performance. After thousands of experiments, a final composition was derived of fused magnetite (Fe3O4) promoted by alumina and potash. Despite the passage of a century, it is striking that most modern ammonia plants use catalysts largely based on the formula defined in the original patent.
Theory of the conventional iron catalyst
An ammonia synthesis catalyst must activate hydrogen and nitrogen to allow the formation of ammonia. The process must not form stable intermediates such as metal hydrides or nitrides that would stop the reaction proceeding. Formation of other stable species that are not catalytically active will inhibit the reaction. In modern plants, the main problem is oxygenates such as CO, CO2 and water.
The active form of the iron in ammonia synthesis is metallic α-Fe. Single crystal studies, molecular modelling and other techniques have been used to understand the nature of the active site, to identify the catalytic mechanism and the rate-determining step.
The mechanism of catalytic ammonia synthesis proceeds via the adsorption of dinitrogen onto the catalyst surface. This is followed by dissociation to adsorbed nitrogen atoms, which is the rate determining step. Direct dissociative adsorption of H2 to hydrogen atoms proceeds, followed by the stepwise addition of hydrogen atoms to the nitrogen atoms to sequentially form adsorbed NH, NH2 and NH3 species, and finally desorption of the NH3 product.
Fig 1: Iron single crystal structures
Surface science studies of the activity of iron single crystals have shown significant structure dependence for the key reaction steps of ammonia synthesis. More open structures such as the Fe(111) structure illustrated in Fig. 1 gave activities of up to two orders of magnitude higher than more closed structures such as Fe (110). The large activity differences may be due to specific surface configurations of atoms, in particular the so-called C7 sites present in the Fe(111) and Fe(211) phases. However, more recent theoretical studies suggest that the actual active sites in the commercial catalyst may be step edges in the iron crystallites.[6,7] These sites are assumed to promote the rate of dissociation of di-nitrogen to nitrogen atoms.
Fig 2: Ammonia formation versus crystal surface orientation
Promoters in the conventional catalyst
The additional components in the ammonia synthesis catalyst are divided into structural and electronic promoters. Electronic promoters, of which potash (K2O) is the most important, increase catalyst activity by changing the electronic nature of the catalyst surface, specifically increasing the rate of nitrogen dissociation. Structural promoters, of which alumina and calcium oxide are the most important, maintain the active structure of the catalyst, in particular its surface area under normal operating conditions. These non-reducible oxides act at the iron crystal grain boundaries as either intergranular particles or a thin skin on the iron crystal surface to prevent iron sintering.
The conventional industrial catalyst
The original commercial BASF ammonia synthesis catalyst contained between 0.6-3.0% Al2O3 and 0.3-1.5% K2O fused with magnetite ore (Fe3O4). This material was then reduced in the process gas to give the active catalyst.
Fig 3: Conventional ammonia synthesis catalyst manufacture and catalyst
The catalysts are produced in the same way today. Exact ratios of starting materials are mixed together and fused at around 1600°C. The main differences are additional promoters (e.g. CaO) and the exact quantities and ratios of promoter species. The molten material is cooled and broken to the required size range.
Fig 4: Relative activity versus particle diameter for ammonia synthesis catalysts
A major change since the 1980s in synthesis converter design allowed the use of the small, high-activity 1.5 – 3.0 mm size catalyst without pressure drop increase. This was achieved by using either vertical radial flow designs (e.g. Ammonia Casale; Topsøe; Uhde) or KBR’s horizontal converter basket design.
Catalyst activation requires reduction of the iron oxide to metallic iron with hydrogen. Oxygen is removed from the magnetite structure with minimal change in crystallite geometry resulting in a high iron surface area.
Often catalyst reduction takes place in situ, using the normal hydrogen/nitrogen feed gas. However, control of the process is vital; the generated water must be kept to a low level, as it can act as a catalyst poison.
Additionally, disposal of ammoniacal water of reduction is a problem in some flowsheets and the reduction process is time consuming.
Consequently, JM offers pre-reduced and passivated versions of its ammonia synthesis catalysts. When commissioned, the stabilising layer is easily and quickly removed at lower temperature, using much less H2 and producing minimal H2O, as illustrated in Fig. 5.
Fig 5: Hydrogen consumption versus temperature comparison for oxidised and pre-reduced ammonia synthesis catalysts
Improvements to iron-based catalysts
As has been described, conventional ammonia synthesis catalyst is produced by fusion of a naturally occurring magnetite iron ore (Fe3O4) with a number of promoters. Precise control of the level of the promoters is essential to give a consistent chemical composition and therefore consistent performance of the finished ammonia synthesis catalyst.
By using modern analysis techniques it is possible to better understand and further optimise the levels and critical ratios of promoters. This has been fed back into the production process to improve the consistency of promoter levels in the KATALCOJM 35-series catalysts. Figures 6 and 7 demonstrate the improved manufacturing controls which have been achieved in moving KATALCOJM 35-series production onto new
Fig 6: Previous manufacturing assets
Fig 7: New manufacturing assets
manufacturing assets. One advantage of the tighter manufacturing capability is that this has enabled ammonia synthesis catalyst to be manufactured that reduces more quickly and maintains a high activity throughout life. An early reference for JM’s improved ammonia synthesis catalyst has been the ammonia plant operated by Misr Fertilizer Production Company (MOPCO) at Damietta, Egypt.
SYNTHESIS LOOP DESIGN
The synthesis loop pressure is around 185 bar and is based on Uhde’s single converter with three radial beds with internal heat exchangers. The make-up gas enters the loop via a combined chiller upstream of the ammonia separator. Separation temperature is set -10°C.
Downstream of the ammonia converter HP steam is generated and routed to the superheater upstream of the HT-Shift. Purge gas and flash gas from the let down vessel are scrubbed in the ammonia recovery unit. The hydrogen is furthermore recovered in a membrane hydrogen recovery unit and routed back to the syngas compressor suction, as shown in Figs 8 and 9.
Fig 8: MOPCO ammonia loop block diagram
Fig 9: MOPCO ammonia synthesis loop
The Uhde-designed converter has three beds with internal heat exchangers. Features of the design are:
• radial-type catalyst beds for maximum conversion rate, lower recycle gas rate and low pressure drop.
• Indirect cooling by heat exchanger for optimum temperature control
• Full bore closure for easy withdrawal of the internal heat exchanger without catalyst removal and comfortable access for catalyst removal without removal of the cartridge
• access to all catalyst beds without removal of intermediate heat exchanger.
The reduction was carried out under a gradually increasing pressure and circulation rate.
Reduction of Bed 1, charged with pre-reduced catalyst KATALCOJM 35-8 was straight forward with oxidic Beds 2 and 3 KATALCOJM 35-4 occurring smoothly immediately afterwards.
Fig 10: Catalysts in ammonia synthesis converter
AMMONIA SYNTHESIS LOOP PERFORMANCE ASSESSMENT
To assist MOPCO’s operation, Johnson Matthey has developed a suite of tools to characterize the catalyst performance based on actual plant data. Especially for ammonia synthesis, an integrated approach is essential as part of the product gas is recycled, which influences the overall rate of production via the prevailing reaction equilibria. A detailed model of the synthesis loop, including the heat-exchange network, was built in the flowsheet package HYSYSTM. The ammonia synthesis reactor was modelled by means of an in-house proprietary developed extension containing our detailed ammonia synthesis reaction kinetics.
Fig 11: Simplified sketch of the ammonia synthesis loop in HYSYSTM
The operating data from the plant were evaluated by using the developed simulation model in combination with in-house developed data reconciliation techniques. The aim of the integrated approach is to perform a heat and mass balance over the main process units of the ammonia synthesis loop to assess plant operation and characterise catalyst performance.
Monitoring the plant operation using the combination of a flowsheet package with rigorous detailed kinetic models and data reconciliation capabilities has a number of benefits:
• The simulation provides a useful consistency check on plant data. This gives confidence that the plant is really operating as the readings indicate and can help identify any process measurements that are inconsistent and hence show where instrument re-calibration may be needed.
• By using the software with accurate laboratory gas analyses, it is possible that errors in plant instrumentation (e.g. flow, temperature measurements) can be identified.
• Regular checks allow any changes in catalyst performance to be identified and investigated.
• The models employed are based on a detailed mathematical model of the plant, including the overall and atomic mass balances, heat balances, reaction equilibria, and reaction kinetics that would be satisfied by perfect measurements on a plant in absolutely steady operation.
The evaluation indicates that the catalyst performance of the installed KATALCOJM 35-series in the MOPCO Ammonia-1 plant is as expected after two years of operation, as illustrated in Fig. 12. This also shows that the plant is operated excellently with respect to the presence of any poisons in the loop. By assessing regular datasets, the actual catalyst performance can be tracked over time and any deviations (e.g. passing driers,failed internals) can be identified early on.
Fig 12: Actual versus expected performance of MOPCO ammonia synthesis catalyst
With the results from the data reconciliation it can be assessed whether there is any further room for improvement, i.e. can the ammonia make be boosted or the loop efficiency be increased. The main variables to be considered are the individual catalyst bed temperatures. An example of the effect of the individual bed temperatures on the ammonia make is shown in Fig. 13. The plots show a distinct optimum, which would allow the production to be increased by about 3 mtpd.
Typically, ammonia synthesis converters are operated close to the optimum given the plant constraints. Therefore, the actual gain that can be realised by optimising bed temperatures is generally in the order of 0.5%.
Fig 13: Ammonia production as function of individual bed temperatures
The fertilizer project for Misr Fertilizer Production Company (MOPCO), which was engineered, constructed and commissioned by Uhde GmbH, has showed excellent performance by exceeding the design capacities at low energy consumption. This could be realized by a combination of good operational practices and skilled local staff, supported by the use of highly active catalysts.
To optimize operation (ammonia make and/or energy efficiency), Johnson Matthey developed a suite of integrated tools to characterize the catalyst performance based on actual plant data. The integral use of a flowsheet package, rigorous kinetic models and data reconciliation techniques enables a detailed assessment of current performance and study improvement opportunities.
The excellent performance of the new plant was demonstrated by achieving an overall production rate of 105.2% and a reliability of 104.8%.
- Tamaru, K.: ‘The history of the development of ammonia synthesis’. In ‘Catalytic Ammonia Synthesis: Fundamentals and Practice’, Ed. Jennings J.R.. Plenum Press, New York (1991).
- Mittasch, A, in ‘Advances in Catalysis’ Vol 2, p. 83. Academic Press, New York (1949).
- Schlögl, R., in ‘Handbook of Heterogeneous Catalysis’ Eds Ertl G., Knözinger H., Schüth F., and Weitkamp J. Wiley VCH, Weinheim (2008) and references therein; Jennings, J.; Ward, S.A., in ‘Catalyst Handbook’ 2nd edn. Ed. Twigg M., Wolfe Publishing Ltd. London (1989)
- Ertl, G.: ‘Reactions at Surfaces: From Atoms to Complexity’. Nobel Prize lecture (2007), available at https://nobelprize.org/nobel_prizes/chemistry/laureates/2007/ertl-lecture.html.
- Ertl, G.: Catal. Rev. Sci. Eng. 21, 201 (1980).
- Norskov et al.: Surf. Sci. 491, 183 (2001).
- Somorjai, G.A., et al.: J. Catal. 03, 213 (1987).