Inside a dissolved acetylene cylinder: porous mass acetone and the numbers that make it safe

A dissolved acetylene (DA) cylinder is the only compressed-gas package in which the contents system, not just the shell, is the safety device. The steel shell is a low-pressure vessel proved at 60 kgf/cm². What stops an acetylene decomposition from running through the charge is a monolithic porous mass of 90–92 % porosity, filled with acetone in which the acetylene is dissolved. Every figure on this page exists to keep that combination inside the envelope in which Berthelot and Vieille found acetone solutions would not propagate a decomposition.

Acetylene is endothermic: it stores +226.9 kJ/mol of heat of formation, and above roughly 1.4 atm absolute a decomposition into carbon and hydrogen can propagate through the pure gas without any air present (Jones, Scott, Kennedy and Huff, US Bureau of Mines RI 3755, 1944). Compressing it to cylinder pressures is therefore out of the question. The history of the DA cylinder is the story of three findings being combined: Claude and Hess showed in 1897 that acetone dissolves an extraordinary amount of acetylene; Berthelot and Vieille reported in 1897 and 1900 that acetone solutions up to 10 atm were not decomposed by detonators or hot wires, explosion being confined to the gas above the liquid; and Le Chatelier and Janet added the porous mass that stops a decomposition in that gas space from spreading.

As compiled by S. A. Miller (1965), the solvent does four things at once. It dilutes the acetylene into a condensed phase. It provides thermal mass. It decomposes endothermically into methane, carbon monoxide and ethylene, so that the heat released per mole of acetylene decomposed falls from 53.7 kcal in the gas to about 40 kcal at 40 g of acetylene per kg of acetone and about 20 kcal at 20 g/kg. And it rate-limits the whole event: decomposition in the gas space is over in milliseconds, but it can only reach the liquid as fast as heat drives acetylene out of solution, so a cylinder decomposition can take up to 100 seconds rather than exploding. See why acetylene is dissolved in acetone for the solvent chemistry.

The modern mass is calcium silicate. Lime and finely ground silica are mixed into a water slurry with a reinforcing fibre, poured into the shell, and reacted under steam at about 185 °C for one to three days, after which the water is driven off to leave the pore volume (as compiled by S. A. Miller, 1965; the same process is described by current manufacturers). The crystalline calcium silicate hydrates formed under those conditions give the mass both its compressive strength and its dimensional stability, which is why strength and freedom from shrinkage are controlled by the same variable, the degree of hydrothermal cure.

Historically the filler was granular or fibrous: kapok, asbestos discs, balsa wood and kieselguhr in the United States, and an 85:15 charcoal–kieselguhr packing in Britain at 78–80 % porosity. Regulators have since closed the door on those. 49 CFR 173.303(f) requires a homogeneous monolithic mass for UN cylinders to ISO 3807:2013, and CGA C-13 condemns all non-monolithic fillers except balsa wood. Indian law still contemplates loose masses, which is why the Gas Cylinders Rules 2016 Rule 39(2) sets a one-year examination interval for them against two years for monolithic masses.

Miller's eight requirements for a mass, compiled from the industry of his day, remain the best short specification: chemical stability over decades in contact with acetylene, acetone and steel; the highest porosity the law allows; mechanical stability against subsidence, cracks and cavities; resistance to the spread of decomposition whether it enters through the valve or starts at a heated wall; chargeability; dischargeability; retention of the acetone during discharge; and cost. The C.P.I., Safety in the Production and Use of Acetylene (Paris 1968), reduced this to four: fill the container as completely as possible, do not disintegrate, oppose the propagation of decomposition, and carry official approval.

Three of these deserve comment.

Take the acetone solubility data of Hölemann and Hasselmann (Forschungsberichte des Landes Nordrhein-Westfalen, 1952–1959), as compiled by Miller: at 15 °C and 15 atm total pressure, 1 kg of acetone holds 546 g of acetylene. Dissolving that much gas swells the acetone by about 84 %, because each gram of acetylene dissolved adds roughly 1.6 ml to the solution volume. So a solvent charge that occupies 43 % of the shell when it goes in occupies far more of the pore volume when the cylinder is full, and the remaining gas space is the entire margin against hydraulic over-pressure. That margin is tested at 65 °C, the tropical design temperature used identically by IS 7312 cl 20.3 and ISO 3807:2013 Annex D, and type-test cylinders are charged to 105 % of the maximum content to prove it.

The settled pressure follows from the same solubility curve. The C.P.I. (1968) recorded the general rule as an effective pressure not exceeding 15 kg/cm² at 15 °C after thermal balance, with filling racks running up to 25 kg/cm²; Indian law caps it at 16 kgf/cm². Because solubility falls steeply with temperature, the same charge at 30 °C sits at a markedly higher pressure, which is the subject of why cylinder pressure does not show contents.

The shell is welded steel to IS 7312 (at most one longitudinal and two circumferential seams), normalised or stress-relieved above 600 °C after welding, and stretch-tested before the mass is cast with no more than 7.5 % permanent expansion. Seamless shells to IS 7285 are also permitted. In service it is not pressure-tested again: IS 8433:2017 states in its foreword that neither a hydraulic nor a pneumatic test, nor an internal visual inspection, can be carried out once the mass is in. (ISO 10462 does require a hydraulic test before new porous mass is introduced into a shell whose test interval has lapsed.)

It is also not a containment vessel for decomposition. AIGA 022/26 puts deflagration final pressures at 10 to 11 times the initial pressure and detonation peaks at about 50 times. From a working pressure of 16 kgf/cm² those are 160–176 kgf/cm² and about 800 kgf/cm², against a shell proved at 60. ISO 3807:2013 cl 4.2 states the design intent plainly: the mass must prevent propagation of acetylene decomposition within the cylinder. The two cylinder fires with published investigations, in Great Britain in 2010 and Taiwan in 2005, were both traced to the same pair of causes: a settled or voided mass, and too little acetone (in the British case, "too little acetone and/or too much acetylene"). The British cylinder was 75 years old with a charcoal mass, so read-across to a modern monolithic fleet is limited, but the pairing is why the filling inspection and the periodic inspection concentrate on exactly those two things.

The valve is to IS 3224:2021, with the acetylene outlet G5/8-14 TPI left-hand, outlet No. 2, and Rule 37(v) limits valve copper content to 70 %. The protection cap under IS 7312 cl 9.2 is perforated steel at least 3.5 mm thick. Fusible plugs to IS 13497:2017 sit in the bung on both sides of the valve shank and yield at 98–104 °C; the shell test pressure may be reduced from 60 to 53 kgf/cm² when they are fitted. Seamless cylinders carry no plug in the body. The valve page and the fusible plug page cover each in detail.