How much water per kilogram of carbide does an acetylene generator need and why 6 to 8 litres

Every carbide-to-water acetylene generator depends on one number: how many litres of water it admits for each kilogram of calcium carbide. KIGL plants are specified at 6 to 8 litres per kg, never below 6.0, and the reason is not tradition. The reaction between carbide and water is strongly exothermic, and the excess water is the only heat sink the generator has.

Calcium carbide reacts with water to give calcium hydroxide and acetylene, releasing about 31.0 kcal per mole of CaC₂ when calculated from the heats of formation in the Handbook of Chemistry and Physics. For pure carbide that is about 484 kcal per kg. Commercial carbide of 300 l/kg gas yield is roughly 82 % CaC₂ and 12 % CaO, and when the heat of hydrating the lime is included it releases about 429 kcal per kg, consumes about 499 g of water, gives 333 g of acetylene and leaves about 1,166 g of solid residue, as compiled by S. A. Miller (1965).

If none of that heat were removed, the 1,166 g of lime and 333 g of gas would be heated by roughly 730 °C. Long before that, the acetylene itself begins exothermic polymerisation and decomposition, so the rise is self-accelerating. Two experiments anchor the point. Hauser (1954) recorded 580 °C within 34 minutes of treating carbide with an equal weight of water. The US Bureau of Mines (Huff, RI 3794, 1945; Jones, Scott, Kennedy and Spolan, RI 4067, 1947) found that moist air at 70 °C heated 14 to 28 mesh carbide until it glowed at about 1,100 °C. Carbide with too little water is not a slow reaction; it is a hot one.

A generator that feeds water and carbide continuously, as every KIGL automatic plant does, settles at an equilibrium temperature set almost entirely by the water-to-carbide ratio. The thermal balances compiled by Miller for a 300 l/kg carbide with 15 °C feedwater at atmospheric pressure give roughly 56 °C at 10 litres per kg, 64 °C at 8 and 75 °C at 6. The gas leaving a wet generator is in equilibrium with the water, so these temperatures are ceilings the generator cannot exceed on its own, provided the feedwater is at the assumed 15 °C.

Where both water and carbide are fed continuously, the codes compiled by S. A. Miller (1965) specify a maximum generation temperature of 75 °C and also require the slurry to stay fluid enough to drain; for batch-water generators the British Acetylene Association's low-pressure code (1938) asked for 10 litres per kg. Both conditions happen to land at the same number: at 6 litres per kg the slurry is about 18 % solids. That coincidence is why 6 to 1 has survived as the industry floor for more than eighty years. The German regulations of 1923, as revised to 1962 (Engel), ask for 10 litres per kg but allow a reduction where clogging and high temperature are prevented by design; they also cap wet generation at 70 °C, which is the ceiling KIGL uses.

Temperatures are the calculated adiabatic equilibria compiled by Miller from Handbook of Chemistry and Physics data with 15 °C feedwater; a real generator loses some heat and runs a few degrees cooler. The 82 °C row is why the water floor must rise with pressure.

Pressure suppresses evaporative cooling, which is a large part of the heat removal at low ratios. As compiled by S. A. Miller (1965), medium-pressure generators at 1.5 to 2 atm absolute could tolerate only about 6.4 litres per kg for the same temperature limit. A KIGL generator at its maximum of 0.92 kg/cm²g is at about 1.9 atm absolute, inside that band. The published KIGL minimum is 6.0; on that figure a plant running at the top of its pressure range should be nearer 6.4. This is a 7 % change and a decision for the works, which is why this page leaves it to the works rather than stating it as a rule.

The practical reading is simple: when the generator pressure switches are set high, run the water ratio in the upper half of the band.

Clause 8.2.1 of IS 8471:2003 requires not less than 8 litres of water for each kilogram of carbide that can be introduced into the holder, unless water and carbide are introduced continuously in a predetermined ratio. Clause 9.2 adds that the total carbide that could fill the holders shall not exceed 1 kg per 8 litres of water in the generating chamber — and, for a water-to-carbide type generator rather than a carbide-to-water one, 1 kg per 4 litres, with the same continuous-feed exemption.

Both clauses are sizing rules, not operating rates, and that distinction is what makes them look stricter than they are. Each one compares the water the generating chamber holds against the carbide the holder could take if it were filled to every available space. Neither prescribes how much water a batch actually consumes. A plant can therefore satisfy clause 8.2.1 on its chamber sizing and still run a consumption ratio below 8 litres per kilogram, which is why the operating band below does not contradict the standard. The exemption very likely covers the continuous-feed design used in KIGL automatic plants, where a pressure-switched screw feed admits carbide and a temperature-controlled valve admits water. The standard also limits the gas temperature at the generator outlet to 75 °C and any external surface to 80 °C at 130 % of rated capacity.

EIGA Doc 237 §4.2 describes the water-to-carbide ratio as approximately 8 to 1, varying between 7 to 1 and 10 to 1. The American codes of the 1960s (Underwriters' Laboratories standards 207 and 403) set 8.3 to 1 for batch-water generators, which is simply 1 US gallon per pound. Current US practice under 29 CFR 1910.253 does not fix a ratio but caps generator pressure at 15 psig, which KIGL's maximum of 0.92 kg/cm²g (13.1 psig) is below.

Feedwater temperature

The thermal balances assume 15 °C feedwater. Miller's compilation is explicit that the equilibrium temperatures hold unless the feedwater, or recycled make-up plus sludge filtrate, is warmer than that, for instance because it has been used to cool the effluent gas. In an Indian summer, feedwater at 30 °C adds roughly 15 °C to every row of the table. At 6 litres per kg that puts the generator near 90 °C against a 70 °C ceiling, which means the temperature controller is doing all the work rather than the thermodynamics. KIGL's manual permits the recirculation of acetylene-enriched water from the lime pit after settling, using a flameproof diaphragm pump, but that water is warm; where it is used, the ratio should be run high and the inlet temperature watched.

Carbide grade

A 250 l/kg carbide (about 68 % CaC₂) releases roughly 391 kcal per kg and leaves about 1,177 g of residue. The thermal floor stays near 6 to 1, but the slurry becomes about 21 % solids, past the 18 % fluidity limit. With poor carbide the ratio must rise to keep the slurry draining and the agitator unloaded.

Slurry solids

The 6 to 1 rule is really two rules: 6 litres per kg and no more than about 18 % solids in the slurry. A plant that keeps the meter at 6 but lets lime accumulate in the tank has broken the second rule while meeting the first. Laboratory series compiled by Miller (Pulkevich, 1934; Strizhevskii, Zaitseva and Shelechnik, 1951) show gasification time unchanged up to 20 % solids with agitation, but doubling at 30 % and lengthening fivefold at 40 %, with large lumps becoming coated and stalling at 20 % without agitation. That is the quantitative case for the agitator timer and for the slurry drain sensitivity check of 200 to 400 litres between successive drain openings.

There is a reason not to drown the generator. Acetylene dissolved in the discharged sludge is lost, and the loss falls with temperature: about 1 % of the make at 60 °C against 0.5 to 0.6 % at 75 °C (Hölemann et al., Forschungsberichte des Landes Nordrhein-Westfalen No. 790, 1959). Running near the 70 °C ceiling rather than well below it is worth about half a percent of output. That loss still dwarfs gasholder losses, measured at 0.007 % of the make for a 20-minute holder (Hölemann, Hasselmann and Strootman, No. 791, 1959), but it is small beside the 13 % yield spread between fine and lump carbide, which is why carbide grading matters more than any water setting.