Dead legs in acetylene pipework: why a closed pipe just longer than the predetonation distance is the worst geometry

A dead leg is a branch of pipe closed at one end: a capped spare connection, an unused branch, an instrument leg, a drain line with its valve shut. In most gas service it is a nuisance because it traps condensate and air. In acetylene service it has a specific and much more serious failure mode. If its length is a little greater than the predetonation distance, a decomposition that enters it has just enough room to turn into a detonation, which then hits the closed end and reflects. That is the geometry that produces the highest pressures ever measured in acetylene pipe.

When decomposition starts in a pipe it usually begins as a deflagration, a flame front moving slower than sound. In a long pipe it accelerates, and after a certain length it can jump into a detonation, a supersonic front with a sudden step in pressure. The length travelled before that transition is the predetonation distance. It shortens as initial pressure rises, and it depends on bore.

Sargent's measurements in 2.5 cm pipe (paper to the International Acetylene Association, Minneapolis, 1957) are the reference set.

Below about 3.5 atm absolute, a deflagration in 2.5 cm pipe tended to die out within 4 m. In 1.9 cm pipe at 16 atm, the minimum predetonation distance was 90 cm.

Narrower pipe at higher pressure gives shorter distances still. Hölemann, Hasselmann and Dix (as compiled by S. A. Miller, 1965) measured 40 cm in 4 mm pipe at 9 atm absolute and 20 cm from 11 to 16 atm; in 8 mm pipe, 60 cm at 8.25 atm falling to 20 cm at 16–19 atm; and in 16 mm pipe, 120 cm at 21 atm and 60 cm at 23 atm. Kogarko and colleagues (Khim. Prom., 1962) found 3.5–11 m in 10 cm pipe 21 m long at 2.45–5.1 atm, again decreasing with pressure.

These distances are short compared with ordinary plant pipework. At cylinder filling pressures, a closed branch less than a metre long is already in the range where a detonation can form.

In a filling plant, the usual candidates are:

• capped spare connections on charging and discharge manifolds;
• branches left in place after a manifold or drier is moved;
• pressure gauge and transmitter legs;
• drain-back and condensate lines with the valve closed;
• bypass loops around a compressor or regulator that are normally shut.

Each one should be checked against the predetonation distance for the highest pressure it can see. Dead legs cause other problems too. The C.P.I. (Safety in the Production and Use of Acetylene, Paris 1968) warned against blind corners and zones that are difficult to clear, because they hold air after purging. They collect water, which on the high pressure side can form hydrates (hydrate plugging). And they cannot be drained if they sit below the line; IS 15190 (Part 1):2002 clause 4.10 asks for a fall of about 1 mm per metre towards drain points.

Instrument legs are a special case because they cannot be removed. EIGA Doc 123/22 asks for a restrictor of 0.5 mm or a flame arrestor on pressure instruments at medium and high working pressure, and IS 15190 (Part 1):2002 clause 7.7.2(b) gives an orifice of not more than 0.1 mm² as an example of a device to limit leakage if the measuring element ruptures. A fine restriction limits flow into the leg and leakage from it; it is not by itself a flame quench at filling pressures, because the quenching diameter shrinks as pressure rises (see the flashback arrestor page), which is why EIGA offers a flame arrestor as the alternative.