Ammonia and the Fertiliser Industry: the Development of Ammonia at Billingham a History of Technological Innovation from the Early 20Th Century to the Present Day

https://doi.org/10.1595/205651318X696341 Johnson Matthey Technol. Rev., 2018, 62, (1), 32–47

www.technology.matthey.com

Ammonia and the Fertiliser Industry: The Development of Ammonia at Billingham A history of technological innovation from the early 20th century to the present day

By John Brightling for their production, ammonia is the most complex Johnson Matthey, PO Box 1, Belasis Avenue, requiring the highest number of catalytic steps. Billingham, Cleveland TS23 1LB, UK Ammonia is one of the most important chemicals produced globally with approximately 85% being Email: [email protected] used as fertiliser for food production (3). The other 15% of ammonia production is used in diverse industrial applications including explosives It is over 100 years since the Haber-Bosch process and polymers production, as a refrigeration fluid began in 1913 with the world’s first ammonia and a reducing agent in nitrogen oxides (NOx) synthesis plant. It led to the first synthetic fixed emissions control. Ammonia synthesis from nitrogen, of which today over 85% is used to atmospheric nitrogen was made possible in the make fertiliser responsible for feeding around 50% first part of the 20th century by the development of the world’s human population. With a growing of the Haber-Bosch process. It remains the only population and rising living standards worldwide, chemical breakthrough recognised by two Nobel the need to obtain reliable, economic supplies prizes for chemistry, awarded to Fritz Haber of this vital plant nutrient for crop growth is as in 1918 (4) and to Carl Bosch in 1931 (5). The important as ever. This article details the historic development of ammonia synthesis directly background to the discovery and development of a addressed “The Wheat Problem” as foretold by Sir process “of greater fundamental importance to the William Crookes in 1898 (6) whereby a shortage modern world than the airplane, nuclear energy, of available reserves (of wheat) would only allow spaceflight or television” (1, 2). It covers the the world’s population to continue to expand to role of the Billingham, UK, site in developing the about two billion which would be reached around process up to the present day. The technology was 1930. Thus, in the early 20th century, the need to pioneered in Germany and developed commercially increase food production led to the development by BASF. In 1998 ICI’s catalyst business, now of the fertiliser industry. Johnson Matthey, acquired BASF’s catalytic Today, the global value of ammonia production expertise in this application and now Johnson is estimated to be over US$100 billion, with Matthey is a world-leading supplier of catalyst and the largest individual plants being capable of technology for ammonia production globally. producing 3300 metric tonnes per day (mtpd) or 3640 short tonnes per day (stpd) (7). To achieve 1. Introduction this scale many improvements have been made over the last 100 years in both process and Ammonia is the second most produced industrial catalyst technology. chemical worldwide. Of the four chemicals, After describing historical aspects of the original ammonia, methanol, hydrogen and carbon ammonia technology development by Haber, monoxide that rely on similar syngas processes Bosch et al. in Germany, and the background to

32 © 2018 Johnson Matthey https://doi.org/10.1595/205651318X696341 Johnson Matthey Technol. Rev., 2018, 62, (1) the requirement for efficiency improvements, this 450 paper uses perspectives from Billingham, UK, to Birkeland-Eyde electric arc method 400 describe some of the technological contributions 350 that came from there in the development of 3

NH 300 ammonia production. –1 Cyanamid method 250 200 2. The Growing Need for Nitrogen 150 Haber-Bosch synthesis In just over 100 years the ammonia production 100 Steam reforming natural gas Energy, GJ mt Energy, industry has grown massively and continues to do 50 } so to feed the ever expanding world population. 0 1910 2010 The development of the remarkable iron catalyst by Alwin Mittasch (8) and the technology for the Year synthesis of ammonia from nitrogen and hydrogen Fig. 1. Historical efficiencies of ammonia process by Fritz Haber and Carl Bosch led to BASF starting technologies to operate the world’s first ammonia synthesis plant in 1913. Researchers estimate that about half consumption can only be reduced marginally, if at of today’s food supply is dependent on the nitrogen all, for the most efficient modern plants. originating from ammonia-based fertilisers (9). Worldwide ammonia production is largely based on Between now and 2050, while the world population modifications of the Haber-Bosch process in which will grow by 30%, the demand for agricultural NH is synthesised from a 3:1 volume mixture of goods will rise by 70% and demand for meat by 3 H2:N2 at elevated temperature and pressure in the 200% (10). This is linked with fundamental shifts presence of an iron catalyst. All the nitrogen used in the demand curve for food, especially caused is obtained from the air and the hydrogen may be by population growth, rising affluence leading to obtained by one of the following processes: changes in diet in many countries and in some • Steam reforming of natural gas or other light regions increasing use of food crops to produce hydrocarbons (natural gas liquids, liquefied fuel. The environmental, human health and climatic petroleum gas or naphtha) aspects of ammonia and fertilisers in the growth • Partial oxidation of heavy fuel oil or coal. scenarios have been reviewed elsewhere (11, 12). In ammonia production technology the type of Ammonia production technology has and feedstock plays a significant role in the amount continues to advance under the competitive of energy that is consumed and carbon dioxide challenges in the industry that demands an ever (CO2) produced. About 70% of global ammonia more energy efficient process, with lower emissions production is based on steam reforming concepts that can operate with high reliability for extended using natural gas, with the use of steam reforming periods between shutdowns. There have been of natural gas considered the best available dramatic increases in environmental performance technology from the point of view of energy use and energy efficiency over the last 100 years, but and CO2 emissions, Table I (14). The use of coal with modern steam reforming processes energy and fuel oil are predominately restricted to China, utilisation is nearing the theoretical minimum (13) which exhibits a strong divergence in the ammonia (Figure 1) and looking forward, specific energy feedstock versus the rest of the world. China

Table I Comparative Energy and CO2 Emissions of Different Ammonia Processes and Feedstocks

Energy, CO2 emissions, Energy source Process –1 –1 GJ t NH3 tonnes t NH3 Natural gas Steam reforming 28 1.6

Naphtha Steam reforming 35 2.5

Heavy fuel oil Partial oxidation 38 3.0

Coal Partial oxidation 42 3.8

33 © 2018 Johnson Matthey https://doi.org/10.1595/205651318X696341 Johnson Matthey Technol. Rev., 2018, 62, (1) accounts for 95% of global coal-based ammonia and Carl Bosch on the technology (5) the ammonia capacity with around 80% of the plants in China synthesis process came to Billingham, UK, in the being coal-based. early 20th century. An ammonia factory being The production of ammonia is a very energy located at Billingham, UK, grew out of the needs of demanding process, the energy use of the World War I when the British government needed to steam reforming process is about 28–35 GJ per develop technology to produce synthetic ammonia –1 tonne ammonia (GJ t NH3). Figure 2 shows the for producing explosives. Billingham was chosen theoretical, practical and operating level energy partly for its proximity to a then-new North Tees efficiencies for ammonia plants based on steam electricity generating station nearby; although reforming. Energy efficiencies vary widely for later developments to the process required less ammonia plants currently in operation due to age, electric power than had been assumed. It is worth feedstock, energy costs and utility constraints. noting that even before the plant was begun the Most plants operate well above the practical possibility for post-war use for fertiliser production minimum energy consumption with the best was recognised. This was recorded in a report by performers (top quartile) ranged between 28 and the Chemical Society in 1916: –1 –1 33 GJ t NH3 and an average efficiency of 37 GJ t NH3. It has been estimated that if all plants worldwide “With some foresight a plant erected were to achieve the efficiency of the best plants, primarily for a military purpose might be energy consumption could fall by 20–25% (15). easily adapted in peace time to agricultural A feature of the industry is that most plants are objects” (16). being continually reviewed for improvements and revamp ideas can be subsequently implemented However by the time the plant (known as the that improve efficiency. Government Nitrate Factory) was completed, World War I was over. The site was put up for sale in 1919 3. Technology Development at (Figure 3), and was purchased by Brunner Mond & Co Ltd (16) who converted it to make ammonia- Billingham, UK based fertilisers. The company was set up as a Following pioneering work by Fritz Haber on the subsidiary called Synthetic Ammonia and Nitrates process (4), Alwin Mittasch on the catalyst (8), Ltd. This became part of ICI in December 1926,

Fig. 2. Energy requirements for Energy loss to Operating ammonia plants Most plants inefficient equipment, level operate in this poor design, limited energy region heat recovery and other factors

28.0 GJ mt–1 NH3 Energy loss to process Practical irreversibility, non- minimum standard conditions energy and byproducts

18.0 GJ mt–1 NH3 Theoretical minimum energy Energy based on ideal chemical reactions, 100% yield, standard state and irreversibility

Fig. 4. Original flowsheet for ammonia production at Billingham

Fig. 3. Advertisement for the Government sale of Billingham Nitrate Plant, November 1919

when ICI was formed from the merger of Brunner Mond, Nobel Explosives, the United Alkali Company and the British Dyestuffs Corporation.

3.1 The Coke Oven Process of Syngas Production

The ammonia plants built at Billingham in the 1920s and 1930s employed the classic Haber-Bosch process based on coke, the same as the production technology used in Oppau, Germany. The first Billingham plant was a 24 mtpd (26 stpd) unit that made its first ammonia in December 1924. The original process is shown in Figure 4. The first stages of gas production were at atmospheric pressure. Alternate streams of steam Fig. 5. Bank of parallel high pressure ammonia and then air were fed into gas generators containing converters, Billingham hot coke to make ‘water-gas’ (hydrogen-rich) and producer gas (nitrogen-rich). These streams were purified using iron oxides to remove hydrogen Using this technology the rise in output from the sulfide and a shift converter to convert most of the site is shown in Figure 6. carbon monoxide to CO2 and H2. The ‘catalysed gas’ As well as scale improvement there were was compressed in reciprocating compressors. CO2 improvements in effectiveness. In 1929, A. H. was removed by counter-current scrubbing with Cowap, Chief Engineer, noted: “a striking feature circulating water and the scrubbed gas was further is an ever increasing rapidity of work. The first compressed, washed with copper liquor to remove large unit No. 3 Unit cost £5¼ million pounds residual CO and CO2 and then fed as make-up and was completed in 27 months (of which 7 gas to the synthesis loop which contained a large months was a labour stoppage for a coal strike). number of parallel converter vessels (Figure 5). No. 4 and No. 5 units cost £11 million pounds and

800 Year Plant developments 1924 No. 2 unit 600 1926 No. 2 unit extension 400 1928 No. 3 unit

200 1929 No. 4 and No. 5 units

Ammonia, mtpd 1932 Extension 0 1941 Process improvement 1920 1930 1940 1950 1960 1951 Converter internals Year redesign

Fig. 6. Ammonia output at Billingham from 1924 to 1951 and the plant developments that coincide with these have been completed in 2 months” (16). Despite Plants, was held in 1955. This meeting became an improvements, by the late 1950s increasing costs important event organised annually to improve of coal and the intrinsic inefficiency of syngas the safety performance of the ammonia industry. generation from coke had made this process It continues accomplishing these objectives by uncompetitive. sharing information on incidents, safety practices, plant performance and technology improvements, 3.2 Partial Oxidation and Plant Safety with the 62nd meeting of AIChE Safety in Ammonia Plants and Related Facilities Symposium being held The first step to improve process efficiency from in 2017 (18). coke-oven syngas production was utilisation of higher pressure oil gasification units, a Texaco 3.3 Steam Reforming of Light gasification unit at Billingham for heavy fuel oil Naphthas was later converted for naphtha feed. Syngas was produced at 30 bar (440 psi) pressure by reaction of Steam reforming of hydrocarbons provides the the hydrocarbon with steam and a limited supply of most economic source of hydrogen gas for ammonia oxygen at 1500°C (2732°F). The partial oxidation synthesis. The general steam reforming reaction is process reduced both the capital and operating shown in Equation (i): costs of low pressure gas generation, eliminated CnH2n+2 + nH2O → nCO + (2n+1)H2 (i) the need for low pressure compression and offered greater feedstock flexibility. The principle The reaction was known to proceed at 700–800°C disadvantage of the process was its requirement (1292–1470°F) over a promoted and supported for an air separation plant to supply oxygen. In nickel catalyst. ICI was amongst pioneers in these early days the challenges for safe operations methane steam reforming and commercial units and engineering of these air separation units (ASU) had been built at Billingham in 1936 to reform were significant. propane/butane byproducts of hydrogenation of In 1959 at Billingham’s partial oxidation coal as part of synthetic hydrocarbons production plant a serious explosion occurred during the (Oil Works). This reforming process was operating commissioning of the ASU which led to three at atmospheric pressure. fatalities (17). This incident resulted in a long In the 1950s natural gas was not available in delay in the partial oxidation plant achieving the UK (discovery and exploitation of North Sea beneficial operation by which time steam reforming gas was still some 15–20 years distant), however technology development had advanced sufficiently increasing quantities of light distillate hydrocarbons to make it a more competitive route for syngas (naphtha) were available at falling prices. Sulfur production. free naphthas had been successfully reformed by Within the industry frequent explosions in oxygen the catalyst research group at Billingham in 1938 plants encouraged engineers to meet and share at atmospheric pressure. What was needed was the information. The very first symposium to discuss development of the process to operate at higher safety in air and ammonia plants, called Safe Design pressures to avoid compression costs. The world’s and Operation of Low Temperature Air Separation first pressurised steam naphtha reforming process

Air Steam Product gas to final purification

Steam Steam

Liquid naphtha

Vapouriser Reforming Boiler Boiler CO2 furnace removal

Hydro- Secondary Shift desulfuriser reformer reactor

Fig. 7. Two stage (primary-secondary) reforming (21)

was designed at Billingham and was brought into (200 psi), the reformed gas, containing 10–12% commercial operation at Heysham, UK, in 1962 (19). CH4, was collected in headers near the ground and The main problems were adequate desulfurisation passed to the air injection burner in the secondary of the feed, and suppression of carbon deposition reformer. After secondary reforming were waste on the reforming catalyst without the use of heat recovery, two stage CO shift, further heat excessive steam ratios. Desulfurisation of the feed recovery, cooling and CO2 removal. The process was addressed by development of feed purification was rapidly adopted and by the mid-1960s over technology involving hydrogenation catalysts 100 steam reforming process licences had been (nickel-molybdenum, cobalt-molybdenum) along sold from Billingham to the following reputed with zinc oxide absorbents capable of reducing engineering contractor licensors: Power Gas sulfur to very low levels. The problem of carbon Corporation (later Davy Power Gas, now Johnson formation was solved by the development of new Matthey), Foster Wheeler (now AMEC), Selas, M. types of alkalised catalysts (20). W. Kellogg (now KBR), Friedrich Uhde GmbH (now Due to equilibrium considerations, to achieve a thyssenkrupp Industrial Solutions GmbH) and low methane slip a temperature of around 1000°C Humphreys & Glasgow (now Jacobs). (1830°F) is required, however the metallurgical limit for a ten-year life of the available tube materials was a design exit temperature of 800°C (1470°F). To overcome this constraint, the new steam reforming process adopted two reforming stages as shown in Figure 7 (21). Now familiar to us as primary and secondary reformers, these unit operations are still present in nearly all ammonia plants.

3.4 Steam Reforming Modernisation

Having developed a viable steam reforming process, the syngas units at Billingham were modernised with four pressured naphtha steam reforming units built in 1962–1963 (Figure 8). Each unit included a primary (tubular) reformer Fig 8. First pressured naphtha steam reforming with 4″ (100 mm) internal diameter tubes and a units for ammonia reaction length of 20 ft (6 m). Operating at 14 bar

The new steam reforming front end occupied Early design memos for the Billingham plants in an area of 14,160 m2 (3.5 acres) – a little less the 1920s had discussed the relative economics than 10% of the area occupied by the coke based of reciprocating and centrifugal compression. processes that it replaced. Using space freed up by They showed that the relative efficiency of rotary the reformers, improvements to the gas purification compressors for the later compression stages would and compression were introduced. The existing be low except at high throughputs (Figure 9). This

CO2 removal process employed water washing low efficiency was one of the reasons for excluding at 55 bar (798 psi) and consumed significant centrifugal compressors for all but the low energy leading to high capital and operating costs. pressure stages and this reasoning still prevailed

Chemical absorbents with higher capacity for CO2 at the beginning of the 1960s; however it was now removal had become available, such as the Benfield challenged. A second plant for ICI Severnside, with process (potassium carbonate) and Vetrocoke a capacity of 545 mtpd (600 stpd), led to what (arsenious oxide), and these were adopted on was described as “possibly the most important different Billingham plants in the early 1960s. event in the history of the development” of the

These processes achieved CO2 slips of <0.1% dry. single stream ammonia plant (23). Figure 10 (24) showing the increase in output of ammonia 3.5 Development of Single Stream converters from 1930 to the mid-1960s illustrates the revolution in scale taking place. By then the Plants normal ammonia unit size was already 600 mtpd As can be seen with the four reformers above, in (662 stpd) capacity and plants with lower capacity contrast to a modern single stream plant, in the than this were being regarded as small. early 1960s many units of process equipment In late 1962, a meeting was held in which Ron (such as compressors and synthesis reactors) still Smith, Vice President of Operations at M. W. Kellogg, had to be used in parallel. A ‘single stream’ concept opined that the capacity of ammonia plants was emerged for new ammonia plants and Billingham bound to increase and queried why the synthesis engineers designed and engineered a 360 mtpd loop pressure could not be reduced from 325 bar (397 stpd) ‘single stream’ plant. Commissioned at (4700 psi) to 150 bar (2200 psi), thus removing Severnside, UK, in 1963, it was, at that time, the the need for reciprocating compressors. Although largest single stream ammonia plant in the world. not successful for that plant bid soon afterwards The plant used the steam naphtha reforming discussions began on how a large plant could be process, a hot potassium carbonate based CO2 built – and the plant design was established which removal system, a copper liquor CO removal quickly became a new technology era that came system and had two high temperature shift (HTS) beds, however parallel reciprocating compressors driven by electric motors were used for synthesis gas compression. 80 Although more active copper-based catalysts were Reciprocating compressors known to be able to accomplish the shift reaction at ~250°C (482°F), they were very sensitive to 70 poisoning by sulfur and could not be used with syngas made from coke. The virtually sulfur free syngas obtained from steam reforming allowed Rotary compressors 60 these copper-based shift catalysts to be used, achieving an equilibrium CO conversion of ~0.2%

dry. ICI developed its own robust and reactive Cu % Stage efficiency, 50 catalyst for this application (22). The combined low level of residual carbon oxides from CO2 removal and CO shift was now low enough that they could be made inert by methanation before the synthesis 40 0 200 400 600 800 1000 loop. As a result of all of the improvements Ammonia output, mtpd considered so far – steam reforming, shift, CO2 removal and methanation – by the mid-1960s it Fig. 9. Mid-1960s final stage compressor was possible to carry out all operations in single efficiencies (%)vs . ammonia output (mtpd) (24) stream reactors.

1000

800

600

400

Ammonia output, mtpd 200

0 1930 1940 1950 1960 1970 Fig. 11. Three M. W. Kellogg reformers Year

Fig. 10. The increase in output (mtpd) of ammonia In keeping with its status as an operator, designer, converters from 1930 to the 1960s (24) technology licensor and catalyst manufacturer, ICI continued to develop its own technology. Two to dominate the industry with a capacity of ‘1000 Billingham designed ammonia plants, constructed stpd’ (900 mtpd). in Kanpur, India, in 1969, featured the first In January 1964, M. W. Kellogg was awarded a application of ICI’s single nozzle secondary burner contract for two ‘1000 stpd’ (900 mtpd) plants to (Figure 12); the forerunner of a design used in be built at Billingham. The design incorporated a many ICI (and subsequently Johnson Matthey) number of important features: designed plants using autothermal reforming • The steam naphtha reforming process at 31 bar technology. (450 psi) pressure • A loop pressure of 131 bar (1900 psi) allowing the use of centrifugal compressors • Improved plant efficiency by recovering heat to generate 103.5 bar (1500 psi) high pressure steam superheated to 450°C (850°F) for use on steam turbine drives. The steam was generated at a higher pressure than that required by the process, so energy was recovered by expanding the steam through turbines to the pressure level required by the process. This greatly enhanced process efficiency. Within a few weeks a third plant was announced, they were the largest plants built at that time.

3.6 M. W. Kellogg Ammonia Units Fig. 12. Single nozzle secondary burner As the M. W. Kellogg plants incorporated the steam naphtha reforming process, Billingham engineers 3.7 Use of Natural Gas Feedstock worked closely with their counterparts from M. W. Kellogg in the design of the reformers, shown in In the 1970s Billingham’s ammonia plants Figure 11. As by the mid-1960s exploration was changed from naphtha feeds to run on the newly ongoing for North Sea gas this was considered commercialised natural gas from the North Sea, and a feature of the reforming process was that however the favourable gas contract was on an the plants could be readily converted to lighter interruptible supply basis, meaning that with short hydrocarbons. notice the feedstock could be cut when demand

39 © 2018 Johnson Matthey https://doi.org/10.1595/205651318X696341 Johnson Matthey Technol. Rev., 2018, 62, (1) for natural gas was high. If natural gas supply was for 1000 mtpd (1100 stpd), this was commissioned interrupted the plants were configured to switch in 1978 and was able to achieve a throughput of feedstock on-line to liquefied petroleum gas (LPG) about 1125 mtpd, (1240 stpd) without significant propane feedstock (which was stored locally in modification. underground salt caverns), bringing a demand for catalysts that could cope with feedstock 3.8 The Ammonia V Process flexibility. This brought new requirements for a catalyst with lower potash and higher activity in Ammonia technology continued to develop and order to optimise the reformer for this feedstock. Billingham-based engineers were tasked with the By the end of the decade there were two light design of a fifth ammonia plant for Billingham potash catalysts: 25-3 (1.6% K2O) for natural (Ammonia V or ‘AMV’). Economic considerations gas feeds and 46-9 (2.2% K2O) for LPG feeds. By meant that capital cost had to be reduced whilst the end of the 1970s, ICI Katalco had a product improving plant efficiency. Although market range very similar to the present: 57-series non conditions in the early 1980s meant that the potash, 25-series light potash, and 46-series plant was never built at Billingham, the designs naphtha catalyst. By this point the catalyst beds for Ammonia V evolved into the AMV process. were operating at temperatures up to 1000°C The first AMV design was commissioned at (1832°F) and 35.6 bar (516 psi), primarily due to Courtright, Canada, in August 1985 (Figure 14), improvements in metallurgy. producing 1120 mptd (1234 stpd) at a total energy In the 1970s, it was recognised that appropriately requirement of 29 GJ per metric tonne (lower formulated low-temperature shift (LTS) catalysts heating value, LHV). Ammonia production was could be self-guarding not only in regard to sulfur, achieved 43 hours after feed gas introduction, but also towards chloride. It was also recognised believed to be a record at that time (25). The AMV that the benefits in terms of shift activity and process also featured a low pressure synthesis bed life accruing from the use of fresh LTS loop operating at about 85 bar (1230 psi) featuring catalyst outweighed the cost savings realised by a new cobalt-promoted high-activity ammonia reusing discharged LTS catalyst. All LTS catalysts synthesis catalyst (KATALCOTM 74-1) which had subsequently developed by ICI and Johnson been developed specifically for the project. Matthey were therefore optimised to maximise As a highly efficient process operating with a their self-guarding capability. low steam ratio, the plant was one of the first to These catalyst systems were utilised in the three suffer from byproduct formation and pressure drop M. W. Kellogg ammonia plants and also in the ICI increase due to HTS over reduction. Copper was designed Ammonia IV plant (Figure 13). Designed added to the HTS catalyst formulation to create an

Fig. 13. Ammonia IV (Ammonia 4) plant at Fig. 14. AMV process plant in Courtright, Canada Billingham

40 © 2018 Johnson Matthey https://doi.org/10.1595/205651318X696341 Johnson Matthey Technol. Rev., 2018, 62, (1) over reduction resistant formulation which was first second plant at Severnside made ammonia only 19 installed in 1987. hours after natural gas was first introduced.

3.9 The Leading Concept Ammonia 3.10 Catalyst Developments Process Catalyst developments continued into the 1990s. By the mid-1980s, the two ammonia plants at Figure 16 illustrates the dramatic improvement in Severnside were becoming uncompetitive and a the activity of one particular catalyst which resulted decision had to be made: improve their efficiency, from a combination of on-going development and replace them or close the site. Improving the the incorporation of learning from the development efficiency was thought unfeasible and it was of the technology for LTS catalysts. decided to develop a new process to replace them. A step change occurred in 1997 due to the This led to the leading concept ammonia (LCA) acquisition of the BASF syngas catalyst business by process technology being developed at Billingham ICI’s catalyst business (since acquired by Johnson (Figure 15). Matthey). The acquisition of the BASF activities The LCA process used a combination of allowed the knowledge of two historic companies new equipment, new catalysts and improved to be combined and in this case the best of both construction and procurement techniques. The companies created a new improved LTS catalyst. range of developments included: Methanol is an unwanted byproduct that may be • KATALCO 61-2 (the first low-temperature formed in LTS reactors and is the main volatile hydro-desulfurisation (HDS) catalyst) organic compound (VOC) emitted from ammonia • PURASPECTM 2020 (the first low-temperature production plants. It is formed as a byproduct in sulfur removal absorbent) both high-temperature and low-temperature shift.

• KATALCO 83-1 (the first application of a process Through the 1990s byproduct methanol was an gas heated reformer (GHR), isothermal shift increasing concern for plants as environmental catalyst specifically developed to resist the high emissions came under closer attention. More operating temperature) selective catalysts became available that made

• KATALCO 11-4 (a low-temperature methanation less methanol. BASF previously had low methanol catalyst) LTS catalysts, K3-110 and K3-111, which suffered

• KATALCO 74-1 (a catalyst which could be used from issues relating to physical strength and in an ammonia synthesis loop at 80 bar (1160 poisons resistance. ICI had LTS products with psi) pressure, even lower than in the AMV good strength characteristics, but could not mimic process). the BASF low methanol recipe due to patent The unique process together with extensive protection. The combination of the two businesses automation start-up sequences meant the plants meant that a low methanol, high-strength product were amongst the most automated ever. The could be developed. The results were KATALCO

83-3K, launched in 1997, and KATALCO 83-3X,

1.4

1.2

1.0

0.8 83-3 0.6 83-2 0.4

Relative shift activity Relative 0.2 53-1 52-2 0 1983 1986 1989 1992 1995 1998 Year

Fig. 16. Relative LTS activity of successive generations of the KATALCO catalysts. The Fig. 15. LCA plant in Severnside, UK numbers within the bars refer to the catalyst series

41 © 2018 Johnson Matthey https://doi.org/10.1595/205651318X696341 Johnson Matthey Technol. Rev., 2018, 62, (1) which was launched in 1998 with 90% reduction purchased in 2014. CATACEL SSRTM is a stackable in methanol byproduct formation compared to structural catalytic reactor for the production previous generations of catalyst (Figure 17). of hydrogen from natural gas. It is made from a high-temperature stainless steel foil coated with a reforming catalyst. This structure allows higher 120 heat transfer and can provide significant capacity Traditional 100 increase to reformers or lower pressure drop 83-3 compared to standard pelleted catalysts. 80 83-3S Further developments have been made by shaping the pellets in some of the other reactors 60 which follow the steam reformer in the production of syngas at the front-end of the plant, notably 40 83-3K the HTS and methanator. For example, KATALCO Relative methanol rate Relative 20 83-3X 71-5F (Figure 18) is a shaped 5-lobe pellet HTS catalyst which exhibits lower pressure 0 drop, increased strength and increased voidage. pre-1993 1993 1996 1997 1998 Similarly, for the methanation reactor, KATALCO Year 11-6MC (Figure 19) uses a 4-hole clover leaf shape Fig. 17. Relative LTS selectivity of the KATALCO to provide lower pressure drop with increased bed catalysts, measured by methanol production rate. voidage. The benefit of pressure drop reduction in The numbers above the bars refer to the catalyst the front end varies from plant to plant depending series on the individual process constraints. Generally

3.11 Developments in Catalyst Shape

The effect of shape on reforming catalysts has been recognised for a long time (21). For steam reforming catalysts, the reaction occurs in a very thin layer at the surface of the pellet. Therefore developments focused on techniques to develop the shape, maximising the external surface area of the catalyst pellets whilst at the same time considering the resistance to flow caused by the way the catalyst packs in the tube. The shape of the steam Fig. 18. Shaped HTS catalyst KATALCO 71-5F reforming catalysts evolved from the original cubes (circa 1930s) to Raschig rings (circa 1940s) to ICI Katalco 4-hole (circa 1980s) and finally the current KATALCO QUADRALOBETM shape. At each iteration, for similar sized pellets the activity increased and the pressure drop decreased (20). Increasing the catalyst activity also allowed the reforming reaction to progress at a lower temperature, which meant the tubes were also at a lower temperature as shown by the measured tube wall temperature (TWT). The lower the peak maximum TWT, the longer the tube metallurgy lasts before failure, with a difference of as little as 20°C (68°F) doubling the tube life. Since the 1990s design tools such as finite element analysis have been used to assist with the design and optimisation of catalyst shape. The latest Fig. 19. Shaped methanation catalyst KATALCO development for the steam reforming process is 11-6MC the CATACELTM technology, which Johnson Matthey

42 © 2018 Johnson Matthey https://doi.org/10.1595/205651318X696341 Johnson Matthey Technol. Rev., 2018, 62, (1) pressure drop is welcome and a small increase in Since then tkIS’s Uhde dual-pressure process has energy efficiency can be gained if it is reduced. been implemented in other similar scale plants recently commissioned in regions of the world with 3.12 The Dual-Pressure Process an abundance of low cost natural gas feedstock (Table II). In 1998, ICI and Uhde (now thyssenkrupp Industrial Solutions (tkIS)) formed an alliance in the field of 4. Ammonia Production Today ammonia technology resulting in a variety of new developments, the most public of which was the Figure 22 shows the current plant capacity and dual-pressure ammonia process (28). The resulting year of construction for all operating ammonia 3300 mtpd (3640 stpd) plant was a step change plants. There is a clear progression of increasing in the scale of plant design available and offered plant scale with time. Market needs for individual a reduction of specific production costs through plants will differ, leading to a range in plant economies of scale. These are still being built today capacities. There are however preferred plant sizes as the world’s largest ammonia plants. which have become ‘standard’ in the industry for The key innovation in the Uhde dual-pressure which references and documented plant designs ammonia process was an additional medium- exist. These can be clearly seen in Figure 22 at pressure once-through ammonia synthesis step capacities of 600 mtpd, 1000 mtpd, 1360 mtpd, operating at around 110 bar (1595 psi), connected 1500 mtpd, 2000–2200 mtpd and most recently in series with the conventional high-pressure 3300 mtpd (3640 stpd). It is notable that, as ammonia synthesis loop at around 200 bar (2900 well as being the largest production units in the psi), Figure 20. The first plant based on this world, the emission limits for the new US fertiliser process was the SAFCO IV ammonia plant in Al projects (ammonia and downstream plants) are Jubail, Saudi Arabia, started up in 2006. With a amongst the lowest in the world, with the design capacity of 3300 mtpd (3640 stpd) it was by far levels for emissions of NOx, N2O, CO and volatile the largest ammonia plant worldwide, Figure 21. organic compounds (VOC) being significantly

Off-gas Second ammonia converter

HP- HP- ~210 bar Purge gas recovery steam steam CW First ammonia converter ~110 bar NH3 chillers NH3

LP HP Once-through Ammonia casing casing ammonia converter from HP loop

Note: molecular CW sieves not shown

NH3 NH3 chillers chiller HP- steam

H2O

CW Make up gas NH3 NH3 chiller Ammonia from once-through conversion from front-end

Fig. 20. Schematic of the Uhde process dual-pressure ammonia synthesis section

below current recognised best available techniques (BAT) values (26). In just over 100 years, the nitrogen fertiliser industry based on ammonia production has grown massively (Figure 23). Drivers behind this growth have been, and remain, increasing global population (Figure 24) (9) coupled with increased plant size to achieve better economies of scale. Although the picture is more complex than this (for example, one could ask which came first: fertiliser or population growth?), together this has created demand for increased capacity and increased reliability from that capacity.

5. Conclusion

Over the last century, scientists and engineers have made a significant contribution to the nitrogen industry. Some of these have been based Fig. 21. SAFCO IV Uhde dual-pressure process (Image courtesy of tkIS) at Billingham, UK, whose heritage now resides with Johnson Matthey and the challenge is to continue

Table II 3300 mtpd tkIS Uhde Ammonia Plants with Johnson Matthey Catalyst

Plant Location Capacity, mtpd Start-up year

Saudi Arabian Fertiliser Al Jubail, Saudi Arabia 3300 2006 Company, SAFCO 4

Saudi Arabian Mining Raz Az Zwor, Saudi Arabia 3300 2011 Company, Ma’aden

CF Industries, Donaldsonville, Donaldsonville, LA, USA 3300 2016 Ammonia 6

Saudi Arabian Mining Raz Az Zwor, Saudi Arabia 3300 2016 Company, Ma’aden 2

3500 Fig. 22. Trends of plant 3000 capacity vs. year of construction 3000 2500 2500 Plant capacity, stpd 2000 2000

1500 1500

Plant capacity, mtpd Plant capacity, 1000 1000

500 500

0 0 1957 1967 1977 1987 1997 2007 2017 Year of construction

170 160 150 140 130 120 110 100 90 80 70 60 50 40 Ammonia production, millions of mtpd 30 20 10 0 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 Year Fig. 23. Global Haber-Bosch ammonia production from mid-20th century to the present. Over 99% of fixed nitrogen production today is by the Haber-Bosch process (2) (Copyright The Fertilizer Institute, used with permission)

7 World population 50 6 World population (no Haber-Bosch nitrogen) Average fertiliser input, kg

40 Meat production, kg person 5 % world population fed by Haber-Bosch nitrogen World population, % /

4 Average fertiliser input 30

Meat production N 20 ha –1 –1 World population, billions World yr 2 yr –1 –1 / 10 1

0 0 1900 1950 2000 Year Fig. 24. Demographic drivers for Haber-Bosch nitrogen and its use in fertiliser: “...the lives of around half of humanity are made possible by Haber-Bosch nitrogen” (2, 9) (Copyright The Fertilizer Institute, used with permission)

this legacy and make an equally significant London, UK, 1899, 207 pp contribution to the future of this vital industry. 7. P. Heffer and M. Prud’homme, ‘Global Nitrogen The fundamental ammonia synthesis process and Fertilizer Demand and Supply: Trend, Current catalysts developed by Haber-Bosch and Mittasch Level and Outlook’, 7th International Nitrogen can still be clearly recognised in even the most Initiative Conference, Melbourne, Australia, 4th– modern ammonia plants. However, the process 8th December, 2016 efficiencies and environmental performances have 8. A. Mittasch and W. Frankenburg, Adv. Catal., been dramatically improved over the last 100 1950, 2, 81 years, most particularly in the preparation of the 9. J. W. Erisman, M. A. Sutton, J. Galloway, Z. synthesis gas, benefiting both ammonia production Klimont and W. Winiwarter, Nat. Geosci., 2008, 1, and other syngas-based processes. Because energy (10), 636 utilisation within modern processes is near the 10. N. Alexandratos, ‘World Food and Agriculture to theoretical minimum, specific energy consumption 2030/50: Highlights and Views from Mid-2009’, can be reduced only marginally, if at all. There Expert Meeting on How to Feed the World in 2050, are many future challenges for ammonia and the Rome, Italy, 12th–13th October, 2009 fertiliser industry, which fall outside the scope of 11. W. Winiwarter, J. W. Erisman, J. N. Galloway, Z. this historical overview. Klimont and M. A. Sutton, Climatic Change, 2013, For now, the ammonia industry will be with us 120, (4), 889 more or less in its present form for decades to come 12. H. J. M. Van Grinsven, J. H. J. Spiertz, H. J. (27). The present production capacity for synthetic Westhoek, A. F. Bouwman and J. W. Erisman, J. ammonia of over 175 million metric tonnes per year Agr. Sci., 2014, 152, (S1), 9 will continue to grow at 1–2% every year to satisfy 13. ‘Energy Efficiency and CO Emissions in Ammonia the increasing demands for food and ammonia- 2 Production’, Feeding the Earth, International based intermediates from an increasing number of Fertiliser Industry Association, Paris, France, people enjoying increasing welfare. December, 2009 14. M. Appl, “Ammonia, Methanol, Hydrogen, Carbon Trademarks Monoxide: Modern Production Technologies”, CRU Publishing Ltd, London, UK, 1997, 144 pp KATALCO, PURASPEC, QUADRALOBE, CATACEL and 15. “Tracking Industrial Energy Efficiency and SSR are trademarks of Johnson Matthey. CO2 Emissions”, International Energy Agency/ Organisation for Economic Co-operation and References Development, Paris, France, 2007, 324 pp 16. V. E. Parke, “Billingham – The First Ten Years”, 1. V. Smil, “Enriching the Earth: Fritz Haber, Carl Imperial Chemical Industries, Billingham, UK, Bosch, and the Transformation of World Food 1957, 110 pp Production”, Massachusetts Institute of Technology, Cambridge, Massachusetts, 2001, 358 pp 17. W. D. Matthews and G. G. Owen, ‘Safety Aspects of Reconstructed ICI Tonnage Oxygen Plant’, in 2. H. Vroomen, ‘The History of Ammonia to 2012’, “Ammonia Plant Safety (and Related Facilities)”, The Fertilizer Institute, Washington, DC, USA, 19th November, 2013 Vol. 5, American Institute of Chemical Engineers, New York, USA, 1963 3. P. Heffer and M. Prud’homme, ‘Fertiliser Outlook 2013–2017’, 81st IFA Annual Conference, Chicago, 18. 62nd Annual Safety in Ammonia Plants and Related USA, 20th–22nd May, 2013 Facilities Symposium, New York, USA, 10th–14th September, 2017 4. ‘The Nobel Prize in Chemistry 1918 – Fritz Haber’, “Nobel Prizes and Laureates”, The Nobel 19. United Nations Interregional Seminar on the Foundation, Stockholm, Sweden, 1918 Production of Fertilisers, Kiev, Ukrainian Soviet Socialist Republic, 24th August–11th September, 5. ‘The Nobel Prize in Chemistry 1931 – Carl 1965, “Fertilizer Production, Technology and Bosch and Friedrich Bergius’, “Nobel Prizes and Use: Papers Presented at the United Nations Laureates”, The Nobel Foundation, Stockholm, Interregional Seminar on the Production of Sweden, 1931 Fertilisers”, United Nations, New York, USA, 1968 6. W. Crookes, “The Wheat Problem: Based on 20. C. Murkin and J. Brightling, Johnson Matthey Remarks Made in the Presidential Address to the Technol. Rev., 2016, 60, (4), 263 British Association at Bristol in 1898: Revised with an Answer to Various Critics”, John Murray, 21. “Materials Technology in Steam Reforming

Processes”, Proceedings of the Materials Haber-Bosch to Current Times’, Proceeding Technology Symposium, Billingham, UK, 21st– 747, International Fertiliser Society, Colchester, 22nd October, 1964, ed. C. Edeleanu, Pergamon UK, 2014 Press Ltd, Oxford, UK, 1966 22. P. Davies, A. J. Hall and D. A. Dowden, ICI Ltd, Further Reading ‘Catalysts of High Activity at Low Temperature’, British Patent Appl., 1968/1,131,631 T. Hager, “The Alchemy of Air”, Harmony Books, New 23. R. H. Multhaup and G. P. Eschenbrenner, York, USA, 2008 “Technology’s Harvest: Feeding a Growing World J. Korkhaus and M. Bachtler, ‘The Ammonia Process Population”, Gulf Publishing Co, Houston, Texas, – A Challenge for Materials, Fabrication and USA, 1996 Design of the Components’, 58th Annual Safety in 24. P. W. Reynolds, ‘Manufacture of Ammonia’, Ammonia Plants and Related Facilities Symposium, Proceeding 89, International Fertiliser Society, Frankfurt, Germany, 25th–29th August, 2013, Vol. Colchester, UK, 1965, 27 pp 54, American Institute of Chemical Engineers, 25. W. K. Taylor and A. Pinto, Proc. Safety Prog., New York, USA, 2013, pp. 167–180 1987, 6, (2), 106 M. Cousins and J. Brightling, ‘Make More From 26. K. Ruthardt, ‘Environmental Constraints on New Less’, Nitrogen + Syngas, 2017, 345, (January- Plant Construction in the USA’, Proceedings 743, International Fertiliser Society, Colchester, UK, February), 52 2014, 24 pp J. Larsen, D. Lippmann and C. W. Hooper, ‘A New Process 27. J. G. Reuvers, J. R. Brightling and D. T. Sheldon, for Large-Capacity Ammonia Plants’, Nitrogen & ‘Ammonia Technology Development from Methanol, 2001, 253, (September-October), 41

The Author

John Brightling is Ammonia Commercial Manager at Johnson Matthey Process Technologies. He obtained his BSc (Hons) in mechanical engineering from the University of Leicester, UK, and has worked in the chemical industry for 30 years. Initially working at ICI with a variety of roles covering plant design, operation and maintenance; for the past 18 years he has worked in the catalyst business with responsibilities for sales, marketing, technical service and product development for the ammonia market. He is a member of the American Institute of Chemical Engineers (AIChE) and has served as Chair of the Ammonia Safety Committee.