What Wet Steam Costs the Equipment Downstream

Why this matters

Wet steam does not announce itself. It shows up as a capacity complaint on a coil that is the right size, years before it shows up as an eroded valve seat or a thinned elbow, and the two are almost never connected by the people who see them. The shop replaces the coil, then rebuilds the control valve, then argues about the coil selection, and nobody has yet looked at the one number that explains all of it.

There is a hazard in this subject that outruns the capacity question. Entrained water in a steam line is what condensate-induced water hammer is made of, and a slug of water accelerated to steam velocity can part a fitting and release live steam into an occupied space. Warm-up on a cold steam line is done by opening the supply valve slowly, in stages, with the drip legs confirmed draining first and every person clear of the line and its fittings for the duration. Do not stand at a joint, a flange, or a valve on a line that is being brought up, and never trace a suspected steam leak with a hand.

Wetness costs two different currencies

Which one you pay depends on what limits the device, and mixing them up is the reason field arithmetic on this subject so often fails to close.

Where the device is fed by the pound, wetness subtracts latent heat pound for pound. A humidifier, a culinary steam injection, a jacket fed through a metered valve: what arrives is a mixture, and the useful heat in a pound of that mixture is the dryness fraction times the latent heat. At 10 psig, from the steam tables, latent heat is about 952 Btu per pound. Steam that is 11 percent wet delivers about 847 Btu per pound of mixture, an 11 percent loss, and 11 percent is the number because the coil or the process is paid in latent heat only. Divide that same 105 Btu against the total enthalpy above 32 F, which at 10 psig is about 1,160 Btu per pound, and you get 9 percent. That denominator is the right one for boiler fuel accounting and the wrong one for equipment output. Say which basis you are on, in the same sentence as the percentage, or somebody will carry the wrong one into a selection.

Where the device is limited by its own surface, wetness costs you surface and temperature instead. A steam coil at a fixed supply pressure does not care whether the water at the tube wall arrived as water or condensed there. What it cares about is that the extra liquid has to drain, that a two-phase mixture drops far more pressure through the same pipe than dry vapor does, and that a lower pressure at the coil is a lower saturation temperature and therefore a smaller temperature difference to work across.

The case below is the second kind, and it is the more common one in the field.

The call

Air handler steam coil, nominal 400,000 Btu per hour, on a 10 psig low-pressure system. Complaint was that discharge temperature fell short in cold weather. The coil had been replaced the previous year on the same complaint.

Measured at the unit, with the system at design conditions: airflow about 8,000 cfm, entering air 42 F, leaving air 83 F.

Sensible capacity from those numbers uses the standard-air constant of 1.08, which is derived at 0.075 pounds per cubic foot and a specific heat of 0.24 Btu per pound per F; the building is near sea level so that constant holds, and at real elevation it falls roughly with density. So 1.08 times 8,000 cfm times 41 F is about 354,000 Btu per hour.

Design leaving air for this coil at 42 F entering is 88 F, which is 1.08 times 8,000 times 46, about 397,000 Btu per hour. Call it the rating. The coil was delivering roughly 89 percent of it, an 11 percent shortfall.

What got eliminated

The coil. Fin surface clean, air-side pressure drop consistent with a clean coil, airflow within a few percent of design. A dirty or bypassed coil was the obvious first candidate and the evidence did not support it.

The trap. Discharging on a normal cycle, condensate line temperature consistent with condensate and flash rather than live steam blowing through. Waterlogging was checked and was not present at the time of the visit.

The station. Downstream gauge at the pressure reducing station read a steady 10 psig, which is what the coil was selected on.

At that point every component reads correct and the coil is 11 percent short.

The two gauges

There is a second gauge, at the coil's own supply connection, and it read 4 psig.

Six psi lost between the station and the coil, across about 60 feet of branch and its fittings, on a system whose entire operating pressure is 10 psig. That is the finding, and one calculation explains it.

The branch is 1.5 in schedule 40, inside area about 0.0141 square feet. Steam demand at the coil's rating is about 400,000 divided by 952, roughly 420 pounds per hour. At 10 psig saturated steam occupies about 16.3 cubic feet per pound from the steam tables, so 420 pounds per hour is about 114 cubic feet per minute, and through 0.0141 square feet that is roughly 8,100 feet per minute.

Saturated steam mains are commonly designed in the 4,000 to 6,000 feet per minute band. This branch runs at about a third again over the top of that band, and one number now explains three separate observations:

• At 8,100 feet per minute the flow is carrying its water as entrained mist rather than a stream in the bottom of the pipe, so nothing about the takeoff geometry or the drip legs excludes it. A sibling article covers why only inertial separation reaches that water.
• A two-phase mixture at that velocity drops far more pressure per foot than dry vapor would, which is where the 6 psi went.
• Droplet impingement erosion climbs steeply with velocity, considerably faster than in proportion, so the same overspeed that caused the first two is quietly machining the trim and the outside radius of the first elbow.

What the pressure loss alone accounts for

Saturation temperature at 10 psig is about 239 F. At 4 psig it is about 224 F. The coil is working across 15 F less temperature difference than it was selected for.

To turn that into capacity, use the coil's own capture fraction, which holds as long as airflow and surface do not change and this coil's did not. At design the coil took the air from 42 F to 88 F while the steam sat at 239 F: it captured 46 of the 197 degrees available, leaving 151, so it holds roughly 77 percent of the available difference in reserve on the leaving side.

Apply the same fraction at 224 F. Available difference is 224 minus 42, or 182 degrees. Seventy-seven percent of 182 is about 140 degrees held in reserve, so leaving air lands near 84 F, an air-side rise of 42 degrees against 46 at design. That is 1.08 times 8,000 times 42, roughly 367,000 Btu per hour, or about 92 percent of the rating.

So the pressure loss alone accounts for roughly 8 points of the 11 point shortfall. Measured leaving air was 83 F, one degree below what the pressure loss predicts.

That residual is worth naming rather than rounding away. Two things sit in it. The entrained water arriving in the tubes adds to the condensate the coil has to drain, and a coil draining more liquid than its trap was selected for holds a little more water in the bottom tubes, which is surface removed from service. The thicker condensate film on a wet-fed tube adds some steam-side resistance too, though on a finned coil the air side carries most of the resistance, so that term is the smaller of the two. Neither is separable with field instruments, and neither needs to be: both are fixed by the same work.

The bill that arrives later

Pull the control valve on a system like this and you find the seat wire-drawn, with the wear concentrated where the flow is fastest. Check wall thickness on the outside radius of the first elbow downstream of the takeoff with an ultrasonic thickness gauge, which requires surface preparation and a trained operator and is a measurement worth buying rather than guessing at, and you find it thinner than the inside radius by a margin that a dry system would not produce in decades.

That is the ordering worth carrying out of this article. The capacity complaint arrives first, because it is proportional and it is felt on the first cold day. The erosion arrives later, because impingement wear compounds with velocity and takes years to show, and it arrives as a separate work order with a separate cause written on it. A shop that answers the first complaint by replacing the coil buys the second one too.

Fixing it, and what fixing it moves

Three items, and they are the same three the one velocity number produced.

Upsize the branch. At 2 in schedule 40, area about 0.0233 square feet, the same 114 cubic feet per minute runs at roughly 4,900 feet per minute, inside the band. That alone recovers most of the 6 psi and most of the 8 points of capacity.

Put a separator ahead of the coil's control valve, selected on the coil's actual flow range from the manufacturer's capacity data rather than on the line size, with its own trap. That removes the entrained water that the branch geometry cannot.

Fix the upstream drip legs and their traps, because the branch is only as dry as what feeds it.

Note what this does not fix. If the boiler is priming and putting carryover into the header, everything above is downstream treatment of an upstream problem, and the water chemistry articles own that. Check the boiler's water level control and its surface condition before spending on separators, because a separator on a priming boiler is a filter on a broken pipe.

Getting an actual dryness number

Nothing in the case above measured steam quality directly, and that is deliberate: there is no hand tool for it. If a number is genuinely required, for a contract acceptance or a warranty argument, it is measured with a calorimeter, and the type depends on the pressure. A throttling calorimeter works only where the sample, throttled to atmosphere, ends up superheated, which needs enough pressure and enough dryness to begin with; below that, a separating calorimeter or a combination instrument is used. Both are laboratory-grade sampling procedures on a live steam line, run by people set up for them, and they belong in a specification rather than in a service van.

For service work, the inference chain in this article is the practical method: confirm the pressure at the device rather than at the station, compute the velocity in the line that feeds it, and treat a line running well over the design band as a wet line until something proves otherwise.

References

• Saturated steam tables from any engineering handbook, for the latent heat, total enthalpy, specific volume, and saturation temperature values used throughout
• Coil and control valve manufacturer selection data, which is the authority for rated capacity, design entering conditions, and acceptable inlet quality
• 29 CFR 1910.147, the OSHA general industry energy control standard, for isolating stored pressure energy before any line is opened for resizing or for a separator installation
• See related: What a Steam Separator Removes and Why; What a Steam Coil Does and Why It Freezes; What a Boiler Does to the Water in It