Calcium carbide storage: why an open drum can lose 20–40 % of its gas in two days

Calcium carbide is a stable solid with one enemy: water in any form, including the humidity in the air. The published numbers on what that costs are stark. Across the whole journey from the furnace tap to the dispatch bay, carbide loses 2–3 % of its gas-making capacity. Left for two days in an open container, 7–25 mm carbide loses 20–40 %. And the air space in an unblanketed drum can build up to 26 % acetylene, deep inside the flammable range, from nothing but ambient moisture. Storage is therefore both the largest yield loss a plant can control and a real explosion hazard, and the two are fixed by the same discipline.

There is surprisingly little published data on carbide deterioration, and most of it dates from the 1950s. S. A. Miller (1965) compiled the following, which remain the reference figures:

The comparison that matters is the first two rows. The producer's entire process, from tapping through cooling, crushing and screening, costs 2–3 %. Two days of careless handling at the user's end cost ten times that. The mechanism is the same in both cases, hydration of the lump surface, and the difference is simply exposed surface area and time.

The 26 % headspace figure deserves to be on the wall of every carbide store. IS 14814:2023 gives the flammable range of acetylene in air as 2.5–82 %, and the C.P.I. (1968) gives the most explosive mixtures as 7–13 %. A drum standing in a humid Indian store, sealed, is therefore not a container of inert solid; it is a small vessel with a flammable atmosphere above the solid, and it must be opened as one. This is the physical basis of the universal rule that drums are opened with non-sparking tools and never near a flame.

Deterioration happens on the surface. A 50 mm lump has a few per cent of its mass within reach of atmospheric moisture; a 2 mm grain is nearly all surface. This is the same reason the standard yield schedules fall from 300 l/kg at 15–50 mm to 248 l/kg at 1–2 mm in the British BS 642:1931 schedule (see carbide grades, sizes and gas yield), and it means the loss in an open drum is far worse for fines than for lump. The 20–40 % figure was measured on 7–25 mm carbide, a common generator feed size. Finer material in the same conditions can be expected to lose more.

The producers understood this early. German practice as recorded in the 1960s was to cool whole 2-tonne ingots in the tapping bogies for 24 hours to a surface temperature of 200 °C, so that cooling took place with the minimum exposed surface, tip them whole into a gyratory crusher, and then keep the crushed product under a dry inert atmosphere. Forced cooling in rotary drums was tried and largely abandoned because the losses were not wholly avoided and the fines in the product increased. The producer's principle is the user's principle: minimise exposed surface, and blanket or seal the product the moment it is cool.

IS 1040:1987 (since superseded by IS 1040:2024, whose text has not been reviewed here) required carbide to be packed in moisture-proof steel containers, of 50 or 100 kg unless otherwise agreed, one graded size per container, marked with the quality and grade, the size, net and gross mass, the month and year of packing, the manufacturer's name and a lot number, and bearing the words CALCIUM CARBIDE — DANGEROUS, IF NOT KEPT DRY with the danger symbol of IS 1260 (Part 1). The month and year of packing is the only shelf-life information a buyer gets, and it should drive a first-in-first-out rotation in the store.

The standard drum of the 1960s, as compiled by S. A. Miller (1965), was light-gauge sheet steel, 50 or 100 kg, usually made at the carbide works and non-returnable, with a pressed-in lid in the centre of the top that is opened first, before the whole top is cut out, and a growing traffic in 250 kg stainless-steel returnable drums. The C.P.I. recorded drums of 50 to 1,000 kg with screw-on, crimped or tin-welded lids. In every case the design intent is the same: water-tight and air-tight until opened, then emptied completely.

The sampling clauses of every carbide specification make the same point in a different way. All of them, as Miller compiled them, insist that sampling be done quickly under dry conditions and that the sample go into a hermetically sealed dry container of a size chosen so that the solid fills it completely. No headspace means no moist air and no acetylene build-up. IS 1040:1987 Appendix D adds that each package should be sampled within about five minutes, in a closed space, with clean dry tools and never with bare hands.