Comparative Evaluation of Ammonia Synthesis Catalyst Features This article examines the development of the ammonia synthesis loop, from the inception of com- mercial ammonia production to today. Opportunities are explored for cost-effective, commercially viable improvements in the ammonia synthesis process through new catalysts and new loop configurations, Claus J. H. Jacobsen and S vend Erik Nielsen Haldor Tops0e A/S, DK-2800 Lyngby, Denmark Introduction mercially viable process was developed. From 1909-1912, Bosch was responsible for the develop- oday, catalytic ammonia synthesis is consid- ment of both the large-scale high-pressure ammonia ered a mature and highly optimized technolo- synthesis converters and the processes for supplying Tgy. Almost 100 years have passed since Fritz sufficient amounts of a sufficiently pure synthesis gas. Haber initiated his pioneering studies of the equilibri- During that same period, Mittasch discovered the dou- um between dihydrogen, dinitrogen, and ammonia bly promoted kon catalyst and many other active com- (Haber and van Oordt, 1904). These studies were the positions, including ruthenium (Mittasch, 1950), in an platform from which Haber developed his process for impressive research program involving catalytic activ- the continuous production of ammonia from the ele- ity measurements of more than 2,000 catalysts in ments (Haber and Elek, 1910). This process, originally almost 6,500 experiments. By 1922, Mittasch and his patented by BASF (DRP, 1908, 1909), revolutionized team had conducted more than 20,000 experiments. the chemical industry and it remains the cornerstone of When the decision to construct a full-scale ammonia catalytic ammonia synthesis. Haber realized that sever- synthesis plant was made in the summer of 1911, the al elements (kon, osmium, uranium, cerium, man- whole technology was completely new and many diffi- ganese, molybdenum, tungsten) were active catalysts culties still remained to be overcome. In view of this, it for the synthesis and decomposition reactions (Haber, is impressive that in September 1913, less than two 1920-1923). However, it was only through the impres- years after the construction was initiated, the fkst sive technical developments by Carl Bosch (Bosch, ammonia synthesis plant was put into operation in 1933) and Alwin Mittasch (Mittasch, 1950) that a com- Oppau, Germany. Soon after, a new plant was con- AMMONIA TECHNICAL MANUAL 212 2002 structed within eleven months by BASF in Leuna near stability was developed (Jacobsen, 2001). It is dis- Leipzig. In 1937, these two BASF plants still account- cussed if new catalysts can be utilized in further ed for more than 70% of the annual world production improved ammonia synthesis processes. We provide capacity (Slack and James, 1973). cost estimates and consumption figures for different Since these earliest developments, the importance of process schemes utilizing different catalysts. From our catalytic ammonia synthesis has steadily increased. In current understanding of the ammonia synthesis reac- 2000, the annual world capacity for ammonia produc- tion, it is possible to accurately predict the maximally tion exceeded 150-106 metric tons (t) of ammonia and achievable activity of an ammonia synthesis catalyst the ammonia production required around 1.5% of the under specified reaction conditions. Based on this, we total global energy consumption (Appl, 2000). Today, speculate about possible future developments of the about 85% of all ammonia is used for production of ammonia synthesis loop. nitrogen-containing fertilizers and thereby ammonia plays a central role in sustaining the growing popula- Developments of the Ammonia Synthesis tion of the world. It has been estimated that at least two Loop billion of the current global population can only be nourished through provision of proteins available via Refer to Slack and James (1973), Appl (2000), dinitrogen fixation by the Haber-Bosch process (Smil, Topham (1985), Hooper (1991), Vancini (1970), and 1997). Consequently, a strong driving force for contin- Dybkjaer (1995). The developments by BASF stimu- uously improving the technology exists. In the scientif- lated other researchers to develop alternative synthesis ic community, a massive effort has been focused on the schemes, partly to circumvent the elaborate worldwide understanding of the fundamental details of this cat- patent protection of the German developments and alytic reaction. Thus, ammonia synthesis has been the partly to search for improvements. The first successful cradle and testing ground for chemical concepts since plant outside Germany was started in Terni, Italy in the beginning of the 20th century (Somorjai and 1920. Some of the earliest synthesis loops employed Materer, 1994). Despite these efforts, few fundamen- significantly higher pressures (about 750-1,000 bar) tally new catalysts have been discovered and many than the BASF plants, thereby allowing increased con- controversies regarding the detailed energetics of the version per pass and a simplified ammonia separation. elementary reaction steps remain, although important In the U.S., after the less successful "US Nitrate Plant insight has resulted from single crystal studies No. 1" (using a sodium promoted cobalt catalyst) com- (Somorojai and Materer, 1994; Ertl, 1991; Somorjai, missioned in 1918 in Alabama, a new plant was put 1991; Ertl, 1980). However, during the last few years, into operation in Syracuse, NY in 1921. Early develop- our understanding of the ammonia synthesis reaction ments in Europe included the technology by Fauser, has significantly improved. used in a plant operated with coke oven gas from the Bosch already realized that generation of the pure company Mont Cenis. This process operated at a very synthesis gas would be the largest single contributor to low pressure (100 bar) and featured a catalyst based on the total production cost of ammonia (Bosch, 1933). iron cyanide. The ammonia synthesis process in the Although most of the significant improvements of the form originally developed by Haber and Bosch deliv- ammonia synthesis process during the 20th century ered ammonia with an energy consumption of 80-90 have been in the synthesis gas generation (front end), GJ/t. After this pioneering work, the modern history of we focus this article on the improvements of the ammo- industrial ammonia synthesis can largely be catego- nia synthesis loop that have occurred since the first rized in three eras based on the available synthesis cat- plants were constructed. Some of the recent process alysts, synthesis loops, energy consumption figures, developments rely on new catalysts (US, 1987, 1979) and available production capacities. In the classifica- with significantly higher activities than those of the tra- tion outlined below, the trends are illustrated with some ditional, promoted iron catalysts. Recently, a new of the common catalysts and synthesis loop configura- ruthenium catalyst with an unprecedented activity and tions. AMMONIA TECHNICAL MANUAL 213 2002 1965-1985: Integrated plant design utilizing 1930-1965: Small units utilizing internal cooling quench or indirect cooling and iron catalysts and iron catalysts 500-1,200 MTPD, multibed ammonia synthesis 100-500 metric ton per day (MTPD), single bed loops, quench or indirect cooling, Mittasch-type cata- ammonia synthesis loops, internal cooling, Mittasch- lysts, energy consumption between 45 and 32 GJ/t. type catalyst, energy consumption between 80 and 45 During this period, the concept of integrated plant GJ/t. The first ammonia synthesis loops all used axial design was pioneered. This was achieved through con- flow converters. The first ammonia synthesis converter struction of large single-train plants with high degrees built in Haber's laboratory in Karlsruhe already relied of energy integration. Reciprocating compressors were on an internal feed-effluent heat exchanger. However, replaced by centrifugal compressors and most new electrical preheat of the synthesis gas was necessary. plants were based on steam or naphtha reforming at Soon after, Bosch constructed the first autothermal pressures of 15-30 bar. Instead of internally cooled ammonia synthesis converter (pilot scale) by improved converters, quench cooling or indirect cooling was pri- insulation and heat recovery. These features were gen- marily employed in synthesis loops containing more erally maintained in the industrial ammonia synthesis catalyst beds operating around 150-220 bar. Although converters of this period. Industrially, the heat of reac- most converters of this period had several catalyst tion was removed using internal cooling in either coun- beds, this was not a new development since the early tercurrent (such as TVA-type converters) or cocurrent Fauser-Montecatini reactor used more beds with indi- (such as NEC-type converters) flow in cooling tubes. rect cooling. The promoted iron catalyst was still Throughout the period, only Mittasch-type promoted exclusively used for industrial application. However, iron catalysts were employed. Due to the axial flow use of the inherently more active, smaller catalyst par- configuration, larger catalyst particles were used (typi- ticles (down to 1.5-3 mm) was made possible through cally 6-12 mm) resulting in pressure drops of 15-20 the invention of radial flow and horizontal converters. bar around the synthesis loop and significant rate limi- This decreased the pressure drop hi the synthesis loop tations due to mass transport restrictions.