Heavy inorganics: salt, sulphur, nitrogen and phosphorus: Sulphur and sulphuric acid: the workhorse acid

1.1 Most sulphur is now a by-product of cleaning oil and gas, so supply ignores demand

Sulphur reaches an acid plant by three routes. Elemental sulphur is the cleanest feed. Pyrites (iron sulphide) can be roasted in fluid beds at 850-940 C to give a gas of 7-14% sulphur dioxide (SO₂). Copper, zinc and lead smelters also give off SO₂, which they must capture anyway. Kent (2012) shows the shift. Mined sulphur fell from 3.2 million tonnes in 1999 to 0.3 million in 2009. Sulphur recovered from oil and gas rose from 37.5 to 46.0 million tonnes over the same decade.

Recovered sulphur comes from the Claus process. Hydrogen sulphide (H₂S) stripped from sour gas or refinery streams is partly burned to SO₂ in a furnace at 950-1,250 C. The rest of the H₂S then reacts with that SO₂ over a catalyst: 2H2S + SO₂ → 3S + 2H2O. Modern three-stage units recover over 98% of the sulphur, against 85-90% in the late 1980s.

This changes the economics of the raw material. A refinery must get rid of its sulphur at almost any price, because it has little storage and cannot stop desulphurising fuel, so small surpluses cause price collapses (Kent). Spot sulphur fell to the tens of dollars a tonne at Tampa and Vancouver in mid-2001, against $140 and $108 in 1990. The Tampa price reached zero in the first quarter of 2009. For an acid maker, the main raw material is often cheap but always volatile.

India has almost no elemental sulphur of its own. Dryden (1997) notes that nearly all Indian acid plants burn imported sulphur. The exceptions are smelter-gas plants at Khetri, Ghatsila, Udaipur and Alwaye, and one pyrites plant at Amjhore in Bihar.

• Kent (2012), Ch 25 Sulfur and Sulfuric Acid · pdf pp. 1001-1009, 1013
• Dryden (1997), Ch IIA Sulfur and Sulfuric Acid · pdf pp. 65-66

1.2 The contact process burns, converts and absorbs, and double absorption removes the last SO2

The contact process has five steps: make SO₂, clean and cool the gas, convert SO₂ to sulphur trioxide (SO₃) on a catalyst, cool the SO₃ gas, and absorb it (Kent). In a sulphur-burning plant, molten sulphur is sprayed into a furnace with dried air, giving gas at 1,000-1,200 C with 10-12% SO₂. The air is dried first, because water would form corrosive acid mist downstream.

In the converter, SO₂ + 1/2 O2 → SO₃ releases heat and is reversible. Equilibrium favours low temperature, while the rate favours high temperature. Shreve (1984) explains the compromise. Gas enters the first bed of vanadium pentoxide catalyst at about 410-440 C and heats itself as it reacts. In a Monsanto four-pass converter, the first pass took the gas from 410 to 602 C and did 74% of the conversion. After cooling between beds, the next three passes added 18.4, 4.3 and 1.3 points, for 98% overall.

Single absorption stops at 97.5-98% conversion. Double absorption takes the SO₃ out after the second or third bed in an intermediate absorber, reheats the gas and sends it through the last bed or beds. With SO₃ removed, the reaction can move forward again, and overall conversion reaches 99.5-99.8%, or 99.9% with caesium catalyst. Kent calls this the industry standard. From the early 1970s, US rules limited new sulphur-burning plants to 2 kg of SO₂ per tonne of acid.

Absorption is done in acid, not water. SO₃ meeting water makes a fog that is almost impossible to catch. It is absorbed in acid near 98.3% H₂SO₄, where vapour pressures are lowest, with over 99.9% efficiency. Water is added to hold the acid's strength, and surplus acid is drawn off as product.

• Kent (2012), Ch 25 Sulfur and Sulfuric Acid · pdf pp. 1013-1016
• Shreve (1984), Ch 19 Sulfur and Sulfuric Acid · pdf pp. 347-356
• Dryden (1997), Ch IIA Sulfur and Sulfuric Acid · pdf pp. 73-76

1.3 A modern acid plant is also a power station, and much of its margin lies there

Three reactions in the plant release heat: burning sulphur, oxidising SO₂ and forming H₂SO₄. Kent puts the thermal output of a 1,000 t/day sulphur-burning plant at about 63 MW. About 97% of the energy input comes from burning sulphur. Only 3% is electricity, mostly for the main blower.

The arithmetic per tonne is simple. 63 MW for one day is 63 x 86,400 = 5.4 million MJ, or about 5.4 GJ per tonne of acid. Most plants recover 55-60% of this as high-pressure steam at 40-60 bar and 400-480 C, which is roughly 3 GJ per tonne. About 40% is lost as warm water from the acid coolers. Shreve put steam output from large plants at over 1.3 t per tonne of acid in 1984.

The lost 40% sits in acid at 110-120 C, too cool to raise useful steam. Monsanto's heat recovery system, first run at Namhae in South Korea in 1987, runs the absorber at about 200 C in a special stainless steel. It raises 10-bar steam and lifts total recovery to about 90%. Twenty-one were built by 2005 (Kent).

When energy was cheap, plants were simple. Once energy prices rose, the heat became a product, plants were redesigned for cogeneration, and acid costs fell (Shreve). A plant beside a phosphate complex can use its steam; a stand-alone plant may struggle to sell it.

Capital is modest by chemical-industry standards. Shreve gives roughly $25 million for an 1,800 t/day double absorption, sulphur-burning plant in the early 1980s. A plant on smelter or pyrites gas cost three to four times as much, because it needs roasters, dust removal, scrubbing and gas reheating, and it exports much less steam.

• Kent (2012), Ch 25 Sulfur and Sulfuric Acid · pdf pp. 1017-1018
• Shreve (1984), Ch 19 Sulfur and Sulfuric Acid · pdf pp. 346-349, 358
• Brennan (2020), Ch 4 Operating cost estimation, Table 4.4 · pdf p. 110

1.4 Acid follows fertiliser, and India built many small plants beside its customers

World output of new sulphuric acid was 158.7 million tonnes in 1999 and 191.3 million tonnes in 2008 (Kent). China accounted for most of the growth, from 21.6 to 53.8 million tonnes. India rose from 6.1 to 7.9 million tonnes.

One customer dominates. The United States made about 27 million tonnes in 2008, and some 74% went into fertilisers, mainly phosphates. Petroleum refining, which uses the acid as a catalyst for alkylation (joining small hydrocarbons into petrol), is the largest of the other uses. Copper leaching, titanium dioxide, pulp and steel pickling follow. The acid seldom ends up in the product: a phosphate plant turns it into gypsum.

Acid is low in value per tonne and awkward to move, so plants sit near their customers. Dryden explains that in India it proved cheaper to carry sulphur inland and make acid beside the fertiliser plant than to ship acid from a coastal plant. The result in 1993-94 was about 6 million tonnes a year of capacity spread over many dozens of sites. Many were built to a standard design of about 33,000 t a year, which is about 100 t/day. Dryden gives typical capacities of 50-1,000 t/day, while Shreve describes a 2,800 t/day plant started in North Carolina in 1982.

Small plants cost more per tonne to build and run, and they export less steam. But they avoid freight and fit a scattered fertiliser industry. Dryden blamed high Indian chemical costs on small plants running below capacity, costly imported equipment, raw material imports and dear power and transport.

• Kent (2012), Ch 25 Sulfur and Sulfuric Acid · pdf pp. 1010-1011, 1018-1019
• Dryden (1997), Ch IIA Sulfur and Sulfuric Acid · pdf pp. 64, 77-82
• Dryden (1997), Ch II introduction · pdf p. 64
• Shreve (1984), Ch 19 Sulfur and Sulfuric Acid · pdf p. 347

See it in the table

Check your understanding

Why does a double absorption plant reach about 99.7% conversion when a single absorption plant stops near 98%?

Roughly how much heat does a sulphur-burning acid plant release per tonne of acid, using Kent's 63 MW for a 1,000 t/day plant?

Why did Indian acid plants tend to be small and spread out?