Sizing Condensate Return Lines: Why the Water Is Not the Problem

A condensate line is not a water line

The most expensive assumption in a steam system is that the pipe downstream of a steam trap carries water. It does not. It carries water and steam together, and the steam takes almost all of the room.

Condensate arriving at a trap is saturated water — as hot as water can be at that pressure without boiling. The moment the trap discharges it into a lower-pressure line, it is holding more heat than water can hold at the new pressure, and the surplus boils off instantly. That is flash steam. It is not a fault, it is not a leaking trap, and it cannot be designed out. It is simply what happens when hot water sees a pressure drop.

The quantity is set entirely by the two pressures, not by how much condensate you have. Discharging to a vented atmospheric receiver, condensate at 3 kg/cm²(g) flashes about 8% by mass, at 7 kg/cm²(g) about 13%, and at 14 kg/cm²(g) about 19%.

Thirteen per cent of the mass, ninety-nine per cent of the volume

The mass figures make flash steam sound like a rounding error. The volume figures are what actually size the pipe.

Take 1,000 kg/h of condensate at 7 kg/cm²(g) discharging to atmosphere. About 133 kg/h flashes to steam and 867 kg/h stays as water. But steam at atmospheric pressure occupies roughly 1,600 times the volume of water, so those 133 kg/h fill about 222 cubic metres an hour while the 867 kg/h of water occupies under one.

Put another way: the flash is 13% of the mass and 99.6% of the volume. A line sized on the water is undersized by a factor of a few hundred, and the flash has nowhere to go.

This is why a correctly selected trap on a correctly sized process can still leave a plant with condensate backing up, waterlogged heat exchangers and hammering returns. The trap is fine. The pipe after it was sized for the wrong fluid.

Falling common lines

Where the common line falls in the direction of flow, gravity is working with you. The water runs along the bottom of the pipe and the flash travels above it, and the two do not fight each other.

Size these on the flash steam volume at the receiver pressure, and check that the water film has somewhere to run. Because gravity is helping, a falling line can generally take the smaller of two candidate sizes where a selection falls between them.

Lay the line with a consistent fall. A falling line with a sag in it is a falling line with a water seal in it, and a water seal in a two-phase line is where hammering starts.

Rising common lines

A rising line is a different problem, and it is the one that catches people out. Here the trap has to push condensate uphill, and it can only do that using the pressure differential available across it.

Every metre of lift costs roughly 0.1 kg/cm² of that differential. Add the friction loss along the line and any back pressure from other traps discharging into the same header, and a trap with 7 kg/cm² upstream can find itself with very little left to work with.

Rising lines should be sized generously — where a selection falls between two sizes, take the larger. The velocity has to stay low enough that the water is lifted as a body rather than atomised and carried, and the extra bore also reduces the friction loss that was eating the differential in the first place.

Better still, avoid the lift. Bring the trap discharge down to a low-level header wherever the layout allows, and lift the condensate with a pump rather than with trap differential.

The mistakes that actually cause backing up

Sharing one common line between traps at very different pressures. The high-pressure traps flash more, raise the pressure in the shared line, and the low-pressure traps then have to discharge against that. The low-pressure equipment waterlogs first, and it is usually the equipment nobody suspects.

Sizing on the condensate rate from the equipment datasheet. That number is the water. It is the right input for selecting the trap and the wrong input for sizing the pipe after it.

Forgetting that the load on startup is far higher than the running load. A main that behaves perfectly at three in the afternoon can be overwhelmed at six in the morning.

Stall on temperature-controlled equipment. When a control valve throttles a coil to hold a temperature, the pressure inside the coil falls. If it falls below the back pressure in the return line, condensate stops draining entirely — no trap of any size will fix that, because there is no differential left to drain with. The fix is a pump-trap or a vented arrangement, not a bigger trap.

How to size one properly

Start with the condensate load in kg/h, and be honest about whether it is the running load or the startup load.

Work out the flash fraction from the trap pressure and the receiver pressure, then convert the flash to a volume at the receiver pressure. That volume, not the mass, is what the pipe has to pass.

Choose a bore that keeps the two-phase velocity within sensible limits, then decide the final size by whether the line falls or rises — down a size for falling, up a size for rising.

For the liquid drain line up to the trap, a different and simpler rule applies: keep the water below about 1.5 m/s on runs under 10 metres and below about 1.0 m/s on longer ones, with pressure drop under 200 and 100 Pa/m respectively. That line carries only water, so it is the one place the intuition works.