# Heavy inorganics: salt, sulphur, nitrogen and phosphorus: Ammonia: fixing nitrogen with natural gas Source page: https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/p3/ [Understanding Chemicals](https://www.materialinsights.org/learn/chemicals/) / [Module 4: Heavy inorganics: salt, sulphur, nitrogen and phosphorus](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/) Part 3 of 5 ## Ammonia: fixing nitrogen with natural gas Ammonia is the gateway to almost all fixed nitrogen. Making it means building hydrogen from natural gas and steam, cleaning that gas to parts per million, and forcing it to react with nitrogen at high pressure. Energy per tonne has roughly halved since 1950, but gas is still most of the cost. ### 3.1 Haber-Bosch is an equilibrium fight won with pressure, a promoted iron catalyst and a recycle loop N2 + 3H2 → 2NH3 releases heat and turns four molecules of gas into two. By Le Chatelier's principle, which says a system at equilibrium shifts to oppose any change imposed on it, high pressure and low temperature favour ammonia. But the triple bond in nitrogen makes the reaction very slow when cool. The industrial answer, found by Haber at BASF in 1908 and scaled up by Bosch by 1913, is a compromise (Kent). The catalyst is iron with a few per cent of aluminium oxide and about 1% potassium oxide as promoters, which stop it sintering (Shreve). Conditions are 350-550 C and 100-250 bar (Kent); Shreve gives about 20 MPa. Even so, only 20-30% of the gas reacts in one pass through the converter. The loop cools the outlet gas, condenses ammonia by refrigeration, mixes the unreacted gas with fresh make-up gas and recompresses it. Methane and argon enter with the feed and build up, so a purge stream holds them at 10-15% of the loop gas. Modern plants recover hydrogen from the purge instead of burning it, cutting energy use by close to 15% (Shreve). Dryden lists processes from about 100 to 1,000 atm, and notes the trend to lower pressure and more recycle because high-pressure vessels are dear. The decisive change came in the 1960s and 1970s, when centrifugal compressors driven by steam turbines replaced reciprocating machines (Kent, Shreve). They suit very large flows at moderate pressure. They allowed single-train plants of 1,000-1,500 t/day and, with other changes, cut costs by close to half (Shreve). Value chain Where US ammonia went (1981) - **Ammonia** over 19 Mt consumed in the US in 1981; about 80% for fertilisers - **Nitric acid** about 20%; mostly to ammonium nitrate - **Ammonium nitrate** fertiliser and explosives - **Adipic acid, nitrobenzene, TDI** nylon and polyurethane chains - **Urea** about 20%; fertiliser, resins, feed - **DAP and MAP** about 15% - **Ammonium sulphate and direct application** other nitrogen fertilisers - **Fibres and plastics** acrylonitrile (0.48 t NH3 per t), caprolactam Source: Shreve (1984), Ch 18 Nitrogen Industries; Kent (2012), Ch 22 Synthetic Nitrogen Products, Shreve pdf p. 322; Kent pdf p. 908 Common misreading #### The nitrogen is not the expensive part Nitrogen is 78% of the air and costs nothing. The expense is the hydrogen. Kent notes that ammonia makers think of themselves as buyers of natural gas rather than of hydrogen, because the hydrogen plant is built into the ammonia plant. The nitrogen arrives with the air fed to the secondary reformer, where the same air burns some of the gas to supply heat. - Kent (2012), Ch 22 Synthetic Nitrogen Products · pdf pp. 883-884, 902-903 - Shreve (1984), Ch 18 Nitrogen Industries · pdf pp. 322-327 - Dryden (1997), Ch IIE Nitrogen Industries · pdf pp. 130-131 ### 3.2 Most of an ammonia plant is a hydrogen factory: reform, shift, scrub and methanate Natural gas first passes over a hydrodesulphurisation catalyst and an absorbent, because sulphur and chlorine poison every catalyst downstream (Kent). Then comes primary reforming, CH4 + H2O → CO + 3H2, which absorbs 205 kJ per mole. It takes place in nickel-catalyst tubes inside a fired furnace at 700-830 C and 15-40 bar. Steam must stay in excess or carbon deposits on the catalyst. In the secondary reformer, hot compressed air is added and burns part of the gas. The temperature rises above 1,200 C and almost all the remaining methane reacts. This also brings in the nitrogen, so most plants need no air separation unit. The water-gas shift, CO + H2O → CO2 + H2, then turns carbon monoxide into more hydrogen. It runs in two steps. An iron-chromium catalyst at about 400 C brings CO down to about 3%. A copper-zinc catalyst at 200-220 C brings it to 0.2-0.4%. CO2 is now the main impurity. It is scrubbed out with hot potassium carbonate (the Benfield and similar processes), with amines such as activated MDEA, or with physical solvents. Kent gives the energy for this step as over 50,000 kcal per kmol of CO2 in the first large plants of the 1960s, which used MEA, and about 10,000 kcal by 2001. Methanation at 250-300 C converts the last traces of CO and CO2 into methane, leaving about 5 ppm. A turbine-driven compressor then raises the 3:1 hydrogen-nitrogen gas to synthesis pressure. Shreve repeats the saying that an ammonia plant is really a steam plant that happens to make ammonia; it raises about four times as much steam as ammonia by weight. Process flowchart Ammonia from natural gas by steam reforming, per tonne of ammonia Inputs **Natural gas (feed and fuel)**about 30 million BTU, about 32 GJ (late-1990s plant) **Process air**supplies the nitrogen (0.96 t N2 at 85% loop yield) **Steam**for reforming and shift; mostly raised in the plant **Boiler feedwater and cooling water**12.5 t cooling water for the synthesis section Process steps 1 **Desulphurisation** hydrodesulphurisation catalyst and absorbents 2 **Primary reformer** nickel catalyst in fired tubes, 700-830 C, 15-40 bar 3 **Secondary reformer** process air added; over 1,200 C; nearly all methane reacts 4 **High and low temperature shift** Fe-Cr at about 400 C to 3% CO; Cu-Zn at 200-220 C to 0.2-0.4% 5 **CO2 removal** hot potassium carbonate or amine; residual 50-1,000 ppmv 6 **Methanation** 250-300 C; oxides to about 5 ppm 7 **Compression and synthesis loop** iron catalyst, 100-250 bar, 350-550 C, 20-30% per pass; purge holds inerts at 10-15% 8 **Refrigeration** ammonia condensed and separated Outputs **Liquid ammonia**1 t **Carbon dioxide**by-product, sent to urea **Steam**raised in waste heat boilers; some exported **Purge gas**hydrogen recovered, rest burned **Process condensate**1,500-2,000 ppm NH3, stripped and reused Dryden's synthesis-section figures are 850 kWh of power and 0.21 t of H2 per tonne; plant sizes 100-1,500 t/day. Source: Kent (2012), Ch 22 Synthetic Nitrogen Products; Dryden (1997), Ch IIE Nitrogen Industries, Kent pdf pp. 889-906; Dryden pdf pp. 128-129 In the data #### Other feedstocks cost more Kent quotes European industry ratios. Taking natural gas as 1.0, heavy oil needs 1.3 times the energy, 1.4 times the investment and 1.2 times the production cost. Coal needs 1.7, 2.4 and 1.7. Oil and coal are gasified with oxygen, which adds an air separation plant and more gas cleaning. Kent notes that about 60% of China's nitrogen fertiliser was coal-based. - Kent (2012), Ch 22 Synthetic Nitrogen Products · pdf pp. 887-892, 899-902, 905 - Shreve (1984), Ch 18 Nitrogen Industries · pdf p. 325 ### 3.3 Energy per tonne roughly halved in fifty years, yet gas is still 70-85% of the cash cost Shreve gives 55.9 GJ per tonne in 1950 and 26 GJ for the best design by 1982. Plants then operating used 40-45 GJ, and he quotes a theoretical floor of about 21 GJ. Kent gives about 42 million BTU per tonne (about 44 GJ, at 1.055 GJ per million BTU) for a world-scale 1970s plant, about 36 million (38 GJ) after retrofitting, and about 30 million (32 GJ) for plants built in the late 1990s. Roy, writing Kent's fertiliser chapter, says the best gas plants use under 33 GJ per tonne of nitrogen, about 27 GJ per tonne of ammonia. The world average, raised by naphtha, oil and coal plants, is 49-55 GJ per tonne of nitrogen. The gains came from engineering, not new chemistry: centrifugal compressors, better CO2 solvents, purge gas recovery and better heat recovery. Gas still dominates cost. Kent states that natural gas makes up 70% to over 85% of cash production cost as its price moves from $2 to $6 per million BTU. His rule of thumb is that cash cost per tonne is 40-50 times the gas price per million BTU. At $4 that is $160-200 a tonne, of which 30 million BTU of gas at $4 is $120. That is why new capacity moved to gas-rich countries such as Trinidad and the Gulf from the 1970s, and why 30% of North American production was shut by 2002, squeezed by imports and dear gas (Kent). Scale is the other lever. Plants of 1,000-1,500 t/day became standard, and 3,000 t/day plants were being designed in 2002. Kent says doubling the size of a single train cuts the capital-related cost of production by about 20%. Data Energy used per tonne of ammonia over time | Plant or period | Energy per tonne of NH3 | Source | |---|---|---| | Typical plant, 1950 | 55.9 GJ | Shreve | | World-scale plant, 1970s | about 42 million BTU (about 44 GJ) | Kent Ch 22 | | Plants operating, early 1980s | 40-45 GJ | Shreve | | 1970s plant after retrofit | about 36 million BTU (about 38 GJ) | Kent Ch 22 | | Best design available, 1982 | 26-27 GJ | Shreve | | Plants built late 1990s | about 30 million BTU (about 32 GJ) | Kent Ch 22 | | Best gas plants, about 2010 | under 33 GJ per t N (about 27 GJ per t NH3) | Kent Ch 24 | | World average, about 2010 | 49-55 GJ per t N (about 40-45 GJ per t NH3) | Kent Ch 24 | | Theoretical minimum quoted | about 21 GJ | Shreve | Conversions: per t N x 14/17 = per t NH3; 1 million BTU = 1.055 GJ. Source: Shreve (1984), Ch 18 Nitrogen Industries; Kent (2012), Ch 22 Synthetic Nitrogen Products; Kent (2012), Ch 24 Fertilizers and Food Production, Shreve pdf pp. 325, 327-328; Kent pdf pp. 905-906, 970 Case #### India's feedstock patchwork Dryden gives 1989-90 costs of producing ammonia at four Indian plants of 900 t/day: IFFCO Phulpur on naphtha, Rs 3,269 a tonne; NFL Panipat on fuel oil, Rs 4,043; RCF Trombay V on natural gas, Rs 4,133; and FCI Ramagundam on coal, Rs 11,601. The coal plant cost about three times as much per tonne. Dryden's plant list shows the newer 1,350 t/day trains at Hazira, Vijaipur and Aonla running on gas. - Kent (2012), Ch 22 Synthetic Nitrogen Products · pdf pp. 884-887, 905-907 - Kent (2012), Ch 24 Fertilizers and Food Production · pdf p. 970 - Shreve (1984), Ch 18 Nitrogen Industries · pdf pp. 325, 327-328 - Dryden (1997), Ch IIE Nitrogen Industries · pdf pp. 131-136 ### See it in the table [H₂O Water](https://www.materialinsights.org/commodities/water/)[Gas Natural gas](https://www.materialinsights.org/energy/natural_gas/)[Cu Copper](https://www.materialinsights.org/commodities/copper/)[Al Aluminium](https://www.materialinsights.org/commodities/aluminum/)[Zn Zinc](https://www.materialinsights.org/commodities/zinc/)[Ni Nickel](https://www.materialinsights.org/commodities/nickel/)[Cr Chromium](https://www.materialinsights.org/commodities/chromium/)[Cl2 Chlorine](https://www.materialinsights.org/chemicals/chlorine/)[H2 Hydrogen](https://www.materialinsights.org/chemicals/hydrogen/)[O2 Oxygen](https://www.materialinsights.org/chemicals/oxygen/)[N2 Nitrogen gas](https://www.materialinsights.org/chemicals/nitrogen_gas/)[Ar Argon](https://www.materialinsights.org/chemicals/argon/)[NH3 Ammonia](https://www.materialinsights.org/chemicals/ammonia/) ### Check your understanding In the ammonia plant, where does the nitrogen for synthesis usually come from? Air fed to the secondary reformer burns part of the gas for heat and leaves its nitrogen in the synthesis gas, so most steam-reforming plants need no air separation unit. Using Kent's rule of 40-50 times the gas price, what is the cash cost of ammonia when gas costs $3 per million BTU? 40-50 x $3 = $120-150 a tonne. Gas makes up 70% to over 85% of that cash cost. Why does the synthesis loop need a purge? Only 20-30% of the gas reacts per pass, so inert gases entering with the feed accumulate. A purge holds them at about 10-15%; hydrogen is now recovered from it. [← Part 2: Salt: chlor-alkali, hydrochloric acid and soda ash](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/p2/)[Mark read · Part 4: From ammonia and rock to fertiliser: nitric acid, urea, ammonium nitrate and phosphates →](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/p4/) ## Other sections - [Module overview](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/index.md) - [Sulphur and sulphuric acid: the workhorse acid](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/p1/index.md) - [Salt: chlor-alkali, hydrochloric acid and soda ash](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/p2/index.md) - [From ammonia and rock to fertiliser: nitric acid, urea, ammonium nitrate and phosphates](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/p4/index.md) - [Industrial gases, and the heavy inorganics side by side](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/p5/index.md) - [Key terms, quiz and sources](https://www.materialinsights.org/learn/chemicals/heavy-inorganic-chemicals/review/index.md)