A Spring Steam Trap is a thermostatic condensate-discharge valve that uses a spring-loaded bellows or capsule charged with a volatile liquid to open below saturation temperature and snap shut as live steam approaches. Unlike a mechanical float trap that responds to liquid level, this design tracks the saturation curve directly through fluid pressure inside the capsule. The purpose is simple — purge condensate and air without losing steam. Trap manufacturers such as Spirax Sarco quote subcooling discharge of 5-15 °C below saturation, which keeps tracing lines and unit heaters efficient on plant from 0.5 bar up to 32 bar gauge.
Inside the Spring Steam Trap
The working element is a small sealed capsule — usually a thin-wall stainless bellows — partially filled with a water/alcohol mix. Inside the capsule the liquid sits at its own vapour pressure, which always tracks a few degrees below the saturation curve of pure water outside. When cool condensate surrounds the capsule, the internal charge stays liquid, the bellows contracts under spring load, and the valve head pulls clear of the seat. As temperature climbs toward saturation, the charge flashes to vapour, internal pressure rises, the bellows expands, and the head drives onto the seat before live steam can escape. The whole cycle is self-regulating — there is no external sensor, no float, no lever train.
The spring matters more than people realise. Without it, the bellows would behave as a pure pressure-balanced element and could chatter on the seat under fluctuating line pressure. The spring biases the closing point so the trap discharges with a consistent 5-15 °C of subcooling — what we call the discharge temperature offset. If the spring is too stiff, condensate backs up and waterlogs the upstream coil. Too soft, and the trap blows live steam. On a Spirax Sarco BPT13 the offset is factory-set and you do not field-adjust it.
Failures cluster around three things. The bellows fatigues from waterhammer and ruptures — when this happens the trap fails open and you'll hear continuous blow-through at the discharge. The seat erodes from wiredrawing if condensate is dirty, and you see condensate carryover even when the trap is cycling. Or the capsule's internal charge leaks, the trap fails closed, and the upstream line floods. The bellows weld is the usual culprit — it must be a continuous TIG seam with no porosity, because a single pinhole at 10 bar working pressure will empty the charge inside an hour.
Key Components
- Thermostatic capsule (bellows or diaphragm): Sealed stainless element, typically 20-40 mm diameter, charged with a water/alcohol or water/methanol mix. The internal vapour pressure curve sits 5-15 °C below the external saturation curve, which sets the discharge offset.
- Bias spring: Compression spring inside or around the capsule that holds the valve open at cold start and biases the closing point. Typical rate 2-5 N/mm; if it loses temper the trap stops discharging cleanly.
- Valve head and seat: Hardened stainless head, usually 3-6 mm diameter, lapped onto a matching seat. Seat hardness must be ≥ 40 HRC or wiredrawing will groove it within 12 months on dirty steam.
- Body and cap: Forged or cast stainless or bronze body rated to PN40 or higher. The cap is removable so the capsule and seat can be swapped without cutting the trap out of the line — a real benefit on heritage installations.
- Inlet strainer: Integral mesh screen, typically 0.8 mm aperture, that catches scale and pipe scarf. Without it, the seat lasts months instead of years.
Where the Spring Steam Trap Is Used
Spring Steam Traps suit any duty where you need tight discharge close to saturation and you want the trap to also vent air on start-up — they are excellent air vents because the capsule sees the cool air the same way it sees cool condensate. You'll find them on tracing lines, unit heaters, autoclaves, jacketed kettles, and small process steam loads. They are less suited to heavy condensate loads with rapid load swings — there a float-and-thermostatic trap usually wins. Sizing is governed by condensate load in kg/h and differential pressure across the seat, and you need to ensure the upstream line cools enough for the capsule to actually close.
- Food and beverage processing: Steam tracing on chocolate transfer lines at Cadbury Bournville, where a Spirax Sarco BPT13 maintains line temperature without live-steam waste.
- Heritage steam preservation: Drain traps on the steam-heating circuits of preserved Pullman carriages on the Bluebell Railway, where the low condensate load suits a small thermostatic trap.
- Hospital sterilisation: Autoclave jacket and chamber drains on Getinge HS66 sterilisers, where a Spring Steam Trap holds the chamber close to saturation during the sterilisation hold.
- District heating: End-of-main drip traps on the central London steam network operated by Pimlico District Heating Undertaking, where they purge waterlogged sections after long no-flow periods.
- Brewing: Wort kettle jacket condensate removal at Fuller's Griffin Brewery, sized for the moderate condensate rates seen during the boil hold.
- Chemical process: Reactor jacket drains on small batch reactors at AstraZeneca Macclesfield, where the trap doubles as an air vent during start-up warming.
The Formula Behind the Spring Steam Trap
Trap sizing comes down to one equation — the maximum condensate the trap can pass through the seat orifice at the differential pressure across it. At the low end of the typical operating range — say a tracing line with only 5-15 kg/h of condensate at 1 bar differential — almost any small trap will cope and the question becomes one of cycle frequency and air-venting capacity. At nominal duty, you size for roughly 2× the calculated steady-state load to absorb start-up surge. At the high end, near 32 bar with a small orifice, sonic conditions limit the discharge and you need to step up an orifice size or fit a second trap in parallel. The sweet spot for a single BPT13-class trap sits between 1 and 14 bar differential with condensate loads of 50-300 kg/h.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| ṁcond | Condensate mass flow capacity through the trap seat | kg/s | lb/h |
| Kv | Discharge coefficient for the seat geometry, typically 0.6-0.8 for a flat-seat thermostatic trap | dimensionless | dimensionless |
| Aorifice | Cross-sectional area of the seat orifice | m² | in² |
| ρ | Density of subcooled condensate at the trap inlet | kg/m³ | lb/ft³ |
| ΔP | Differential pressure across the trap seat (inlet minus outlet) | Pa | psi |
Worked Example: Spring Steam Trap in a heritage glasshouse heating circuit
you are sizing a Spring Steam Trap on the end-of-main drip pocket of a recommissioned 1923 cast-iron steam-heating circuit at the heritage Kibble Palace glasshouse in Glasgow, where the main runs at 0.7 bar gauge house pressure feeding overhead pipe coils, and the trustees want trap capacity confirmed across a mild-autumn light load, a typical winter nominal load, and a hard-frost peak load before the public reopening.
Given
- Pline = 0.7 bar gauge
- Preturn = 0 bar gauge (atmospheric return)
- ΔP = 70,000 Pa
- Seat orifice diameter = 4.0 mm
- Kv = 0.7 dimensionless
- ρ (condensate at ~110 °C) = 951 kg/m³
Solution
Step 1 — calculate the seat orifice area for the 4.0 mm head:
Step 2 — at nominal winter load with full 0.7 bar differential, compute the trap's maximum discharge capacity:
That is the trap's hydraulic ceiling — it can pass 365 kg/h of subcooled condensate through that 4 mm seat at 0.7 bar drop. The actual discharge will be much less because the bellows cycles open only when the upstream pocket cools below saturation. Step 3 — at the mild-autumn light load with maybe 15 kg/h of condensate at the same 0.7 bar drop, the trap operates at roughly 4% of its hydraulic capacity. You'll see short, infrequent discharges every few minutes, and the trap spends most of its life closed. That is exactly what you want for an end-of-main drip in mild weather.
Step 4 — at the hard-frost peak with the glasshouse pulling maximum heat and roughly 200 kg/h of condensate reaching the drip pocket, the trap sits at 55% of capacity. Still well within margin, but cycle frequency rises and discharge becomes near-continuous.
The 4 mm seat is correctly sized — it has roughly 1.8× margin over the worst-case hard-frost load, which absorbs start-up surge when the cold cast-iron coils first take steam and dump several minutes of condensate at once.
Result
Nominal hydraulic capacity at 0. 7 bar differential is 365 kg/h, against a measured peak demand of around 200 kg/h — a comfortable 1.8× margin. At light autumn load the trap idles at 4% duty with widely spaced discharges, at winter nominal it sits at maybe 25% duty cycling regularly, and at hard-frost peak it climbs to 55% with near-continuous discharge but no backup. If you measure condensate spilling out of the upstream relief or hear waterhammer in the cast-iron coils, the most likely causes are: (1) the inlet strainer is blocked with mill scale dropping ΔP across the screen and starving the seat — pull and clean the 0.8 mm mesh first; (2) the capsule charge has leaked through a porous bellows weld and the trap has failed closed, which you confirm by feeling a cold trap body during a known steaming period; or (3) the discharge line back-pressure has risen above 0.5 bar because the condensate return is waterlogged, collapsing your effective ΔP and cutting trap capacity by a third.
Choosing the Spring Steam Trap: Pros and Cons
The choice between a Spring Steam Trap, a float-and-thermostatic (F&T) trap, and a thermodynamic (TD) disc trap depends on condensate load profile, line pressure, and how much subcooling you can tolerate. The Spring Steam Trap is the all-rounder for low-to-moderate loads where air venting matters. F&T traps win on heavy modulating loads. TD traps win on superheated steam mains where the others would struggle.
| Property | Spring Steam Trap (Balanced Pressure Thermostatic) | Float and Thermostatic (F&T) Trap | Thermodynamic Disc Trap |
|---|---|---|---|
| Maximum working pressure | Up to 32 bar (e.g. Spirax BPT13) | Up to 21 bar typical | Up to 42 bar typical |
| Discharge temperature relative to saturation | 5-15 °C subcooled | At saturation (continuous modulating) | At or near saturation, blast discharge |
| Air venting on start-up | Excellent — primary advantage | Good (separate thermostatic vent built in) | Poor — needs separate air vent |
| Condensate capacity range | Low to moderate, 5-1000 kg/h | Moderate to heavy, 50-15,000 kg/h | Low to moderate, blast cycling |
| Tolerance to waterhammer | Poor — bellows fatigues and ruptures | Moderate — float can be damaged | Excellent — no fragile internals |
| Suitability for superheated steam | Limited — capsule charge can boil dry | Poor — steam can lock the float | Excellent — preferred choice |
| Typical service life on clean steam | 7-10 years (capsule replaceable) | 5-8 years | 3-5 years (disc wears) |
| Installed cost (DN15) | £90-£180 | £200-£450 | £60-£140 |
Frequently Asked Questions About Spring Steam Trap
Continuous discharge with the coil hot almost always means the capsule has lost its charge and the trap has failed open — the bellows is no longer pressurising against the seat. Pinhole porosity in the bellows weld or fatigue cracking from waterhammer cycles are the usual causes. Confirm by isolating the trap, letting the body cool, and listening with a stethoscope on the next steam-up — a healthy capsule snaps shut audibly within a minute or two as saturation approaches. If you hear no shut event, the capsule is dead and needs replacement.
Less commonly, the seat has wiredrawn from years of dirty condensate and the head no longer lands cleanly. You'll see a circular groove on the seat face when you pull the cap.
The deciding question is whether you need condensate removed exactly at saturation or whether 5-15 °C of subcooling is acceptable. On a process kettle where the jacket holds product temperature within a tight band — pharmaceutical manufacture is a good example — the F&T modulates at saturation and gives you better thermal control. On a brewing wort kettle or a heritage glasshouse coil where a few degrees of backed-up condensate doesn't matter, the Spring Steam Trap is cheaper, simpler, and self-venting on start-up.
Rule of thumb: if condensate load swings by more than 4× during the cycle and you have heavy start-up surge, go F&T. Otherwise the spring trap is the better engineering choice.
A hard slam usually means the spring rate is too high for the application or the trap is undersized so it cycles in short, sharp bursts. Each cycle the bellows expands rapidly when the upstream condensate clears and live steam suddenly hits the capsule, slamming the head onto the seat. Over thousands of cycles this fatigues the bellows weld and you'll lose the trap.
The fix is to either step up to the next orifice size so the trap modulates rather than blasts, or to add a small cooling leg of 1-2 m of uninsulated pipe ahead of the trap so the condensate arrives well below saturation and the bellows closes gently.
Generally no, and this is where practitioners get caught out. The capsule charge is a water/alcohol mix with a vapour pressure curve calibrated to saturated steam. On superheated steam the capsule sees temperatures well above its own saturation point, the internal liquid boils completely to vapour, and the bellows can plastically deform or rupture from over-pressurisation.
For superheated mains, fit a thermodynamic disc trap or a bimetallic trap rated for the superheat margin. If you must use a thermostatic trap, install it at the end of a long uninsulated cooling leg so the condensate has dropped well into the saturated region before reaching the capsule.
This is a back-pressure problem, not a trap problem. In winter the condensate return main runs full and cold-water-logged sections raise return pressure. Your effective ΔP across the seat collapses — if line pressure is 1 bar gauge and return pressure climbs from 0 to 0.6 bar, you've lost 60% of your driving pressure and the trap capacity drops with the square root of ΔP, so capacity falls by about 37%.
Check return-line venting and any lift sections. A common culprit on heritage installations is a sagging return main that has filled with debris over decades. Re-grading the line or fitting a pumped condensate receiver usually solves it without changing the trap.
On a cold start the capsule sees ambient air, stays fully contracted, and the valve sits wide open — you get the full seat orifice flow area for air venting. For a 4 mm seat at 0.5 bar driving pressure that's roughly 50-80 m³/h of free air, which clears a typical small process line in under a minute.
Where it falls short is on a large header with a long cold-fill — the trap will pass air, but a single small thermostatic trap cannot evacuate a 100 m main quickly. Fit a dedicated balanced-pressure air vent at the high point of the header instead and let the drip trap handle condensate only.
Yes — the capsule and spring are matched as an assembly to the working pressure band. Higher-pressure traps use a stiffer spring and a different liquid charge so the saturation offset stays in the 5-15 °C window across the operating range. Fitting a 7 bar capsule on a 21 bar line will cause the capsule to overpressurise internally and either rupture the bellows or hold the valve permanently shut as the charge fully vapourises.
Always check the pressure rating stamped on the cap or the capsule itself. On Spirax BPT13 traps the capsule colour-code identifies the pressure range — fit the wrong colour and you've designed in a failure.
References & Further Reading
- Wikipedia contributors. Steam trap. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
