Knowledge / Ship systems

Steam traps and condensate: heat transfer, backpressure and water hammer

Follow steam heating through condensate production, trap differential pressure and flash steam, with examples showing why a healthy trap can still fail to drain a system.

On this page

Steam gives up useful heat when it condenses, but the resulting liquid must leave the intended steam space. A steam trap contributes to that task while limiting the escape of live steam and handling air as its design allows. It does not create pressure to push condensate uphill or through a pressurized return. On ships, where heaters, tracing and return systems operate across changing loads, the whole drainage path matters as much as the trap itself.

Separate steam supply from condensate removal

A heater can receive steam yet perform poorly if condensate occupies heat-transfer area intended for condensing vapour. Air and other non-condensable gases can also interfere with the intended process. A supply-pressure reading upstream of a control valve therefore does not prove that steam reaches every surface, nor that condensate is leaving. Define the boundary from the controlled steam space through the trap to the return destination.

Some systems deliberately control heat transfer by partial flooding and are designed for that duty. Others require prompt drainage and are not intended to operate waterlogged. The same visible level or outlet temperature can therefore have different meanings. Start with the equipment’s design intent before calling every retained liquid volume a fault or assuming that every trap should discharge continuously.

Estimate the condensate load from the heat duty

For a simple steady example, assume a heater transfers 500 kW entirely through condensation and use an illustrative effective latent enthalpy release of 2,000 kJ/kg. The condensate formation rate is 500 kJ/s divided by 2,000 kJ/kg, or 0.250 kg/s, equal to 900 kg/h. This connects thermal duty to a mass load. The enthalpy value is an assumed rounded input; an actual calculation uses steam properties at the relevant pressure and inlet state.

Additional cooling of the condensate, superheat removal, heat loss and non-steady warm-up change the balance. Startup condensate load may differ from the stabilized process load because pipework and metal are being heated. A trap selected only from a nominal steady duty can therefore miss the controlling condition. Conversely, a large catalogue capacity at an unrelated pressure difference is not evidence of adequate capacity in the installed system.

Calculate the pressure actually available for drainage

Spirax Sarco’s stall explanation defines the loss of condensate flow when the steam-space pressure is no greater than the total backpressure imposed on the trap. The pressure downstream of a throttling control valve can fall as heat demand falls. Boiler-header pressure can remain normal while the heater loses the differential needed for drainage. This is a system condition, not necessarily a stuck trap.

Use a consistent pressure datum. In a hypothetical operating state, the steam space is at 2.00 bar absolute and the return connection at 1.50 bar absolute. Assume the downstream liquid lift contributes 0.20 bar and friction at the considered flow contributes 0.10 bar. The remaining differential is 0.20 bar. If steam-space pressure falls to 1.60 bar while the static return conditions remain the same, it cannot overcome the 1.70 bar return-plus-lift requirement even before friction is added. A larger passive trap cannot manufacture that missing pressure.

Distinguish trap principle from installation suitability

The Spirax Sarco trapping overview distinguishes mechanical, thermostatic and thermodynamic families and their different discharge behaviour. Selection depends on the application, not on one family being universally superior. Relevant questions include air handling, permitted condensate backup, changing load, pressure difference and return conditions. The cited catalogue is a general product overview; its ratings are not a vessel-specific selection.

A healthy trap can be unsuitable for a particular installation, and a suitable trap can be impaired by isolation, a blocked strainer or a changed return arrangement. These cases require different corrective reasoning. Record the installed model and its rated conditions, but also demonstrate the actual hydraulic path. Replacing parts without checking the path can reproduce the same symptom after each maintenance cycle.

Recognize flash steam as an energy balance

Hot condensate entering a lower-pressure region may partially vaporize because its incoming enthalpy exceeds that of saturated liquid at the lower pressure. TLV’s flash-steam explanation describes this phenomenon. For an ideal steady adiabatic pressure reduction with negligible kinetic and potential energy change, the flash mass fraction is x = (hin − hf,out)/hfg,out, provided the resulting state lies in the two-phase region. This is different from live steam leaking through a defective trap.

Use a deliberately simplified property example: incoming condensate enthalpy 760 kJ/kg, outlet saturated-liquid enthalpy 505 kJ/kg and outlet latent enthalpy 2,200 kJ/kg. The flash fraction is 255/2,200 = 0.1159, or about 11.6% by mass. For 900 kg/h incoming condensate, the ideal result is approximately 104 kg/h of flash steam and 796 kg/h of remaining liquid. These rounded properties illustrate the calculation; they are not a complete design point or a guarantee of measured discharge.

Do not size a return by liquid mass alone

A modest vapour mass can occupy a large volume compared with the accompanying liquid. The return system may therefore carry two phases even when the upstream equipment mainly produces condensate. Pressure loss, velocities, pooling and interactions between branches depend on more than total kilograms per hour. A liquid-only pipe calculation can underrepresent what occurs after flashing.

Heat loss or subcooling before the pressure reduction can reduce the flash fraction, while mixing streams can change where condensation occurs. The location at which a vapour cloud becomes visible is not a direct measurement of the mass leaking past a trap. Compare the actual pressure and enthalpy states and the test method before classifying a visible discharge as either healthy or defective.

Understand two different water-hammer mechanisms

Liquid accelerated by steam can strike a change in direction or other obstruction, while rapid steam condensation can collapse a pocket and drive a separate pressure transient. TLV’s condensate-return discussion explains several condensation-related patterns where hot vapour meets cooler liquid. The point at which impact is heard may be away from the place that created the unstable mixture. Following sound alone can therefore mislocate the cause.

A simple steady pressure-loss calculation does not predict the peak of such a transient. Nor does a higher pipe pressure rating remove the initiating mechanism. Examine where liquid can accumulate, where hot and cold streams meet, which branches operate intermittently and how startup or shutdown changes the flow. The purpose is to identify the physical source of the transient, not to experiment with rapid valve actions on a live system.

Use several observations to assess a trap

The US Department of Energy steam-trap tip sheet discusses inspection methods including temperature and sound. These methods need operating context. A hot outlet can arise from flash steam or a live-steam leak; a cold trap can reflect no current load, isolation, insufficient driving pressure or a blockage. One surface-temperature reading rarely settles the diagnosis.

Compare the expected operating cycle, upstream and downstream conditions and the actual test response. Record whether the heater was warming up, stable or throttled down. If the test method depends on listening to intermittent discharge, an unfamiliar trap principle can be misread. A maintenance report should distinguish a demonstrated leak, a demonstrated failure to drain and a result that remains inconclusive under the tested condition.

Close the loop with system evidence

After corrective work, verify the required heating duty and drainage behaviour under relevant load conditions, including the conditions that previously caused the symptom. Confirm restored isolation states, return routing and the intended control behaviour. A quiet pipe immediately after repair is useful evidence but does not establish that the low-load stall condition or intermittent mixing problem has been removed.

A useful record contains the heater duty, steam-space and return pressures, elevation basis, trap identity, measured or estimated condensate load, observed cycle and inspection method. Keep a thermal explanation, a hydraulic explanation and the condition of the trap distinct. These three views make it possible to recognize a faulty device without blaming it for a pressure balance that no passive trap can overcome.

Sources