Why Steam Mains Hammer, and How to Stop It

Two different things share one name

Water hammer in a steam system is not one phenomenon. Two quite different mechanisms produce the same bang, and they need different fixes, so it is worth knowing which one you have.

The first is condensate slug pickup. Water lying in the bottom of a main gets caught by steam travelling at thirty metres a second, accelerated into a solid slug, and stopped abruptly by a bend, a valve or a closed end. The energy arrives as a hammer blow on the fitting.

The second is steam bubble collapse. A pocket of steam trapped in sub-cooled condensate condenses almost instantaneously, and the surrounding water rushes in to fill a void that has effectively vanished. This one is more common in flooded heat exchangers and badly drained returns than in mains, and it can be violent enough to split fittings.

Both are avoidable. Neither is a trap problem in the first instance — they are drainage and layout problems.

Startup is when it happens, and here is why

A steam main in service produces a modest, steady trickle of condensate from heat lost through the pipe wall. A steam main warming up from cold produces a torrent, because every kilogram of steel has to be raised from ambient to steam temperature and the heat for that comes from steam condensing on the inside of the pipe.

The arithmetic is worth doing once. A 200 metre DN100 main at 7 kg/cm²(g) holds about 3,500 kg of steel. Warming it from 20 °C to 170 °C in fifteen minutes requires roughly 500 kg/h of condensing steam.

That same main, lagged with 50 mm of insulation, produces about 33 kg/h once it is hot. The startup load is fifteen times the running load.

If the drain points and traps were selected on the running figure — and they very often are, because that is the number on the heat balance — the main simply cannot clear water on startup. It collects, the steam picks it up, and something bangs.

Insulation is not only about fuel

The same 200 metre DN100 main with no lagging at all loses around 1,030 watts per metre and produces roughly 360 kg/h of condensate continuously — more than ten times the insulated figure, and close to the startup load, permanently.

That is a large fuel bill, but it is also a permanent drainage load that most systems were never designed to carry. Stripped or damaged lagging on a section of main is a common and easily missed cause of a return that suddenly starts hammering after years of behaving.

Where drain points belong

Every 30 to 50 metres along a straight horizontal run, and additionally at every low point, at the bottom of every rise, ahead of every isolating and control valve, and at every dead end or blind flange.

Lay the main with a fall of about 1 in 100 in the direction of steam flow, so condensate runs towards the drain points rather than away from them. A main laid dead level is a main with water sitting in it.

Where a main has to rise, drain it immediately before the rise. Steam will not carry condensate uphill; it will simply push it into a pool at the bottom and then pick it up.

The drain pocket matters as much as the spacing

This is the detail that most often defeats an otherwise sound design. Condensate travelling at steam velocity has considerable momentum. A small branch taken off the bottom of a main does not catch it — the water skims straight past the opening and carries on down the pipe.

The pocket has to be large enough for the water to fall into. On mains up to DN100 it should be full bore, the same diameter as the main itself, and at least 100 mm deep. From DN125 to DN200 a DN100 pocket at least 150 mm deep is appropriate. At DN250 and above the pocket should be at least half the main diameter and as deep as the main is wide. These dimensions are set out in EN 45510-1, section 10.1.14.

The depth serves a second purpose: it leaves room below the trap take-off for dirt and scale to settle, rather than being carried straight into the trap and its strainer.

The operational fix and the engineering fix

Operationally, open the main slowly. Most startup hammer is caused by a valve being opened fully on a cold line, which sends steam down it at full velocity before anything has warmed. Cracking the valve open and letting the main come up over several minutes gives the drain points time to work.

An automatic warm-up sequence does the same thing more reliably than an operator with other things to do, and it is worth the small cost on any main long enough to matter.

Engineering-wise: size the traps on the warm-up load rather than the running load, check that the drain pockets are actually full bore, confirm the fall runs the right way, and make sure the discharge lines can pass the flash steam rather than only the water.

When it is not the trap

Traps get replaced far more often than they fail. Before condemning one, check whether the drain pocket is deep enough to catch anything, whether the trap is discharging against back pressure from a shared return, and whether the equipment is stalling — a control valve throttling a coil below the return line pressure stops condensate draining entirely, and no trap can drain against a differential that does not exist.

If hammering starts after years of trouble-free running, look for what changed: lagging removed for maintenance and not replaced, a new user added to an existing return header, or a main extended without adding drain points.