What Superheat Changes About a Steam System

Why this matters

Every convenient thing about steam comes from one relationship: at saturation, pressure sets temperature, so a gauge is a thermometer and a reducing valve is a temperature control. Superheat breaks that relationship on purpose. The steam sits at the same pressure and is now hotter by whatever the superheater put in, and nothing on the outside of the pipe says so.

The result is a system where the instruments still read, the fittings look the same, and four assumptions the shop has carried for years are quietly untrue. This article is which four, and what each costs.

A superheated steam leak is the worst leak in this trade to look for. Saturated steam at least forms a visible plume a short distance out; superheated steam may not form one at all, because it has to shed its superheat before it can condense into anything you can see. Clear the area, then locate with an ultrasonic detector from a distance, or a rag on a long non-conductive pole held from outside the jet path. Never a hand or a sweep of the body.

What superheat is, and the number that names it

Superheat is the temperature a vapor carries above the saturation temperature for the pressure it is at, and degrees of superheat is a subtraction: measured temperature minus saturation temperature at the measured pressure. It is the same idea a tech already uses on a refrigeration suction line, which a sibling article covers. The physics is identical and the consequences are not, because on a steam system saturation temperature is the process control rather than a charging diagnostic.

At 100 psig, saturated steam is about 338 F from the steam tables. A 100 psig header carrying 438 F steam has 100 F of superheat. Two pipes side by side, both gauges reading 100 psig, 100 F apart in temperature and identical in every visible respect.

The one thing superheat genuinely buys

Superheated steam cannot make condensate until its superheat is spent. That is the entire commercial reason it exists.

A saturated main pays for its heat loss by condensing, which is why it needs drip legs, traps on every one, and a warm-up procedure. A superheated main pays for the same loss out of its superheat, dropping in temperature while staying dry, and only begins making condensate once it has come all the way down to saturation. That means dry steam at the far end of a long run, and dry steam is what a turbine has to have, because water droplets at blade velocity erode blades.

So superheat is a transport property. It is not a heating property, and treating it as one is where the trouble starts.

Four things that stop being true

One: the gauge stops telling you the temperature. Every habit built on reading a pressure and knowing what the equipment sees is invalid. A process control that regulates pressure on a superheated supply is not regulating temperature at all, and if the process needs a temperature, it needs a temperature control loop with a temperature sensor.

Two: a flow meter set up on saturated density reads high. Superheated steam at the same pressure is less dense, and a differential-pressure meter infers mass flow from density and differential. The size and the character of that error are both worked below.

Three: a trap sized on running condensate load has no running condensate load. A trap on a superheated line does essentially nothing for months and then has to handle a very large startup slug. Sizing it on the running number gives you a trap sized on zero.

Four: allowable stress and thermal growth belong to the metal temperature, not the saturation temperature. Both are computed at 438 F, not at 338 F, and the difference is not small.

The heat exchanger question, and the condition that governs it

Condensing steam gives film coefficients one to two orders of magnitude higher than a dry gas at the same temperature, and from that it is easy to conclude that a superheated supply forces an exchanger to give up its superheat through a gas-cooling zone that eats surface. Sometimes that is true. Usually it is not, and the thing that decides it is the tube wall temperature.

Where the tube wall sits below the saturation temperature, condensation proceeds on the wall even though the bulk steam is superheated, and the effective coefficient stays near the condensing value. The superheat is given up through a condensing film rather than through a gas film, and the penalty is modest.

Where the tube wall runs above the saturation temperature, which happens in the entry region of an exchanger heating something already hot or heating a gas with a poor coefficient on the other side, there is no condensation there at all. That region is a gas cooler, its coefficient is one to two orders of magnitude lower, and it occupies a share of the surface out of all proportion to the duty it carries.

So ask what a specific exchanger's wall temperature does, not what its bulk steam does. An exchanger heating water from 140 F to 180 F has a wall far below 338 F everywhere and barely notices 100 F of superheat. An exchanger heating air to 300 F on a 100 psig supply may have an entry region above saturation and notices it a great deal.

Traps on a line that makes no condensate

Two requirements, and they pull against each other.

It has to survive running dry at full superheated temperature indefinitely, which rules out any design whose sealing element depends on a fill not rated for it; a balanced-pressure thermostatic capsule filled for saturated service can be damaged above saturation. The trap manufacturer's published superheat rating is the authority for a given model.

It also has to pass the startup load, which is large: every pound of pipe metal goes from ambient to saturation before the line carries superheat, and the Btu come out of condensing steam. That load exists once per startup and then vanishes.

One failure is peculiar to this service. An inverted-bucket trap seals on a water prime, and on a superheated line that prime can boil away during a long dry run. The bucket then sits down and the trap blows through continuously the moment the line does make condensate again, invisibly, on a system where nobody is watching the trap because the line is supposed to be dry.

Worked: a 100 psig header carrying 100 F of superheat

Header at 100 psig, measured 438 F, saturation 338 F, so 100 F of superheat. Feeds a converter that heats building water from 140 F to 180 F, and it is metered.

Enthalpy, and which denominator. At 100 psig, saturated vapor carries about 1,190 Btu per pound and latent heat is about 881. Superheated steam near saturation has a specific heat of roughly 0.5 Btu per pound per F, which varies with pressure and with how far above saturation you are, so treat it as an estimate and use the superheated steam tables for a real selection. On that estimate, 100 F of superheat adds about 50 Btu per pound: 5.7 percent against latent heat, 4.2 percent against the saturated total of 1,190, and 4.0 percent against the superheated total of about 1,240. Three defensible denominators, so the sentence has to name one. Say which one you mean. The latent basis is the one that matters to an exchanger, because latent heat is what it is paid in.

What the converter does with it. Water in at 140 F and out at 180 F, so the tube wall sits far below 338 F across the bundle. Condensation proceeds under a superheated bulk essentially everywhere, the effective coefficient stays near the condensing value, and the 50 Btu per pound comes off through a condensing film. This converter loses very little to superheat, and that 50 is real capacity rather than an accounting artifact. Change the service to a 300 F air heater and that conclusion reverses, for the wall-temperature reason above.

What the meter does with it. Take the steam tables' 3.89 cubic feet per pound at saturation. Scale volume with absolute temperature, converting to Rankine with 459.67 because this is a ratio of two temperatures and the offset does not cancel: 3.89 times 897.67 over 797.67 is about 4.38 cubic feet per pound, so density falls from about 0.257 to about 0.228 pounds per cubic foot, roughly 11 percent. A differential-pressure meter infers mass flow as the square root of density times differential, so one configured for saturated density reads high by the square root of 0.257 over 0.228, about 6 percent.

Now attach the two labels that make that 6 percent usable. Its basis is percent of reading, so it does not shrink at low flow and it is not a fixed count you can subtract. Its character is a fixed systematic offset, not a random spread, so it does not average out over a month of readings and it does not combine in quadrature with anything. If the same offset sits on two meters being differenced it cancels, leaving only that percentage of the difference; against a feedwater meter with its own separate error it does not cancel at all. The ideal-gas volume scaling used above is itself an approximation that gets worse near saturation, so the superheated steam tables, not this ratio, are the authority for a real correction.

What the pipe does with it. Carbon steel expands at roughly 6.5 millionths of an inch per inch per F over this range, a coefficient that itself drifts with the temperature range, so the piping code's expansion table governs a real calculation. From 70 F ambient, a saturated 338 F line grows about 2.1 inches per 100 feet and the same line at 438 F grows about 2.9. Superheat added roughly three quarters of an inch per 100 feet, which has to be absorbed by loops, offsets, or expansion joints and guided so the growth goes where the designer intended. Allowable stress is lower at 438 F as well, which is a design question for ASME B31.1 in the edition adopted by your authority having jurisdiction, not a field call.

Confirming you actually have superheat. Read the gauge, convert to saturation temperature from the steam tables, then measure the pipe. Use a contact probe on metal that is already bare, a valve body, a flange face, a trap body, an existing gap in the lagging, or infrared on a taped or painted target patch of known emissivity, because bare steel is too reflective for an honest infrared reading. Do not open, brush, cut or lift lagging to make a reading: thermal system insulation on steam piping is presumed asbestos-containing under 29 CFR 1926.1101 and 29 CFR 1910.1001 until sampled, and removal belongs to a trained crew. Seat a probe only on bare metal, in gloves rated for that surface and with eye and face protection under 29 CFR 1910.132, remembering a superheated line runs hotter than its gauge implies. Then attach the error character before concluding anything: the surface reading is biased low by the internal film, the wall and any insulation disturbance, and that bias runs one direction only. On bare metal with a seated probe it is a few degrees; through partial lagging it is not bounded at all. So a reading meaningfully above saturation proves the line is superheated. A reading at or below saturation proves nothing, and treating it as proof of saturated service is the mistake this measurement invites.

When the process needs saturated steam back

If a plant distributes superheated and a process needs the saturation link, the superheat is removed with a desuperheater, which injects treated water into the flow and lets it evaporate. Three constraints ride with it and all three are commonly missed.

The water has to be boiler-quality treated water, because everything dissolved in it stays in the steam and lands on the process side. The line downstream needs enough straight length for the spray to fully evaporate, or you have converted a superheat problem into a wet steam problem with the erosion a sibling article covers. And a desuperheater controls to a temperature setpoint, so its sensor has to sit far enough downstream to see mixed flow, which is a longer distance than most installers allow for.

A reducing station is not a desuperheater. Throttling is close enough to constant enthalpy that reduced steam comes out slightly superheated rather than cooler.

References

• Saturated and superheated steam tables from any engineering handbook, which are the authority for specific volume, enthalpy, and specific heat at any pressure and temperature pair used above
• ASME B31.1 Power Piping, in the edition adopted by the authority having jurisdiction, for allowable stress at temperature and for thermal expansion design
• Steam trap manufacturer literature, which is the authority for a specific trap's superheat rating and for whether a given design tolerates prolonged dry operation
• Flow meter manufacturer documentation, for the density basis a meter was configured with and the correction procedure for superheated service
• See related: What Wet Steam Costs the Equipment Downstream; What a Pressure Reducing Station Is Managing