Steam Trap Capacity: A Multiplier Standing In for a Calculation, and Why Warm-Up Is the Hard Case

Two Jobs for One Device

A steam trap has to pass everything that has condensed and hold back everything that has not. The sizing question is which of those two duties the selection is being made against, because they pull in opposite directions.

Steam gives up its latent heat at constant temperature, which is what makes it useful for heating. What it leaves behind is water at that same temperature, and that water has to leave the steam space immediately. Condensate standing in a coil occupies surface that should be transferring heat. Condensate lying in a main and then picked up by steam moving at velocity is what water hammer is made of. A trap that passes too little leaves both conditions in place.

A trap that passes too much passes steam along with the condensate. Live steam through a trap is energy leaving the system without doing any work, and it was raised at the boiler and paid for there.

The two failure modes sit on opposite sides of a single selection, which is unusual. Most equipment sizing has a comfortable direction to err in. A pump a little larger than needed throttles back, a duct a little larger than needed runs quieter. A trap a little larger than needed does not sit quietly, and what it does instead depends on the type of trap it is.

What makes the choice harder is that the condensate load is not one number. There is the rate at which condensate forms once the system is hot and working, and there is the rate while the system is being brought up to temperature from cold. The second can exceed the first several times over, and the ratio between them is a property of the particular thing being heated rather than of steam.

The calculation on the page handles this with a multiplier applied to the running load, which is a convenient shortcut for a quantity that is arrived at separately.

What follows covers what that multiplier stands in for, how the warm-up load is actually calculated and how far it can sit from what a factor of two implies, why the differential pressure the calculator reports decides the answer without entering the formula, and what the capacity ratio does and does not test about a selection.

Calculator Inputs: A Load, a Multiplier, and Three Optional Fields

The field list is five entries long, and only the first two are required.

Field	Imperial unit	Metric unit	Required
Condensate Load	lb/h	kg/h	yes
Startup / Safety Factor	dimensionless	dimensionless	yes
Inlet Steam Pressure	psi	bar	no
Outlet / Back Pressure	psi	bar	no
Selected Trap Capacity	lb/h	kg/h	no

The relations between them are three lines.

Required Capacity     = Condensate Load × Startup Factor
Differential Pressure = Inlet Pressure − Outlet Pressure
Capacity Ratio        = Selected Capacity / Required Capacity

The first two fields give the required capacity on their own. The remaining three work in pairs. The two pressures give the differential, and the selected capacity gives the ratio and the category. Without a selected capacity the calculation returns the required capacity and passes no judgement on any trap, which the page states directly rather than filling the gap with an assumption.

Because the differential is a subtraction of one entered pressure from the other, the two have to be on the same basis. Both gauge or both absolute gives the same difference; one of each does not.

The ratio is placed in one of four bands.

Capacity ratio	Band
below 1.00	UNDERSIZED
1.00 up to 1.15	ACCEPTABLE
1.15 up to 1.50	RECOMMENDED
above 1.50	OVERSIZED

The page calls these illustrative preliminary interpretation bands and states that they are not manufacturer capacity ratings or universal steam trap rules. They are read here in exactly that sense, and they are not attributed to any published document, because no published document sets them.

Two features of the field list are worth naming before going further. The differential pressure is calculated and displayed, but it does not enter the required capacity. That is correct rather than an omission, for reasons the section on differential pressure sets out. And the selected capacity field expects a rated figure at the differential that actually applies, not the largest number on the manufacturer's chart. The field carries helper text saying so.

What the field list does not contain is as informative as what it does. There is no mass of equipment being heated and no warm-up time, which are the two quantities the startup load follows from. There is no trap type, which decides how the trap behaves when it is oversized and when it runs at part load. And there is no lift after the trap and no return header pressure, which are two of the things back pressure is made of.

Warm-Up Is a Different Calculation

The load while a system is warming up follows from the metal being heated and the time allowed to heat it. That is a different calculation from the running load rather than a margin added on top of it, and running it for a real line shows how far apart the two can be.

For pipework, the condensate formed in bringing the metal up to temperature is:

m_condensate = m_metal × c_metal × ΔT / h_fg

m_metal   mass of metal being heated, kg (lb)
          16.1 kg/m (10.8 lb/ft) for DN100 standard wall,
          so a 100 m (328 ft) run is 1,610 kg (3,549 lb)
c_metal   specific heat of the metal, kJ/(kg·K) (BTU/(lb·°F))
          about 0.49 (0.117) for carbon steel; stainless is
          near 0.50 and copper near 0.39
ΔT        rise from starting temperature to saturation
          temperature at the operating pressure, K (°F)
          commonly 100 to 165 K (180 to 297 °F) for mains
          at 1 to 10 bar gauge starting from 20 °C (68 °F)
h_fg      latent heat of vaporisation at the operating
          pressure, kJ/kg (BTU/lb)
          2,258 (971) at 1 bar absolute falling to
          2,014 (866) at 10 bar absolute

The warm-up rate is this mass divided by the warm-up time.

Taking a line of 100 m (328 ft) of DN100 pipe to standard wall, warmed from 20 °C (68 °F) to the saturation temperature at 7 bar absolute (101.5 psia), which is 165 °C (329 °F), a rise of 145 K (261 °F), with a latent heat of 2,066 kJ/kg (888 BTU/lb) at that pressure:

Energy     = 1,610 × 0.49 × 145 = 114,400 kJ (108,400 BTU)
Condensate = 114,400 / 2,066    = 55.4 kg (122 lb)

That mass is fixed. The rate depends entirely on how long the warm-up is allowed to take.

Warm-up time	Warm-up rate
10 minutes	332 kg/h (732 lb/h)
30 minutes	111 kg/h (244 lb/h)
60 minutes	55 kg/h (122 lb/h)
90 minutes	37 kg/h (81 lb/h)

The running load of the same line is a different quantity again. An insulated line of this size loses of the order of 110 W/m (114 BTU/(h·ft)) at this temperature difference, which over 100 m is 11 kW (37,500 BTU/h):

Condensate = 11 / 2,066 × 3,600 = 19.2 kg/h (42 lb/h)

Heat loss depends on the thickness and condition of the insulation, so that figure is an order of magnitude for a line of this size rather than a property of every insulated main. Every ratio below follows from it.

Warm-up time	Warm-up ÷ running
10 minutes	17.3
30 minutes	5.8
60 minutes	2.9
90 minutes	1.9

The expression above is the warm-up of pipework and nothing else. A heat exchanger adds the mass of the product being heated, which for a liquid at ordinary specific heats usually dominates the metal. A jacketed vessel adds its contents. And a plant warming from genuinely cold has cold insulation as well as cold metal, so the loss during warm-up is higher than the steady figure that the running load is built on. The startup load of a real application is a sum of contributions, and the one above is a single term in it.

Per Spirax Sarco guidance on warm-up load and the steam consumption of pipework: the condensate formed while a line is brought up to temperature follows from the mass of metal, the temperature rise and the latent heat, divided by the warm-up time, and it is calculated rather than assumed as a margin.

What the Multiplier Corresponds To

Running the multiplier backwards against the warm-up calculation shows what warm-up time it implicitly allows. The answer is specific to the system, which is the useful part.

The required capacity is the running load times the factor, so a factor of 2.0 selects the trap for twice the running load and no more. For the line above, twice the running load is 38.3 kg/h (85 lb/h), and the warm-up rate falls to that value at a warm-up time of about 87 minutes. A factor of 2.0 therefore covers a warm-up stretched over roughly an hour and a half and nothing faster. At 30 minutes the warm-up rate is 111 kg/h (244 lb/h), so the factor that would cover it is 5.8.

How the ratio comes out depends on the mass of metal per unit of running load. A long distribution main with good insulation gives a high ratio, because there is a great deal of steel and very little steady duty. A heat exchanger with a high thermal output and a moderate mass gives a low one. The same numerical value of the multiplier therefore means different things in the two cases, which is why it cannot be carried from one application to another.

The page presents 2.0 as a value used in some steam applications and notes that larger values may be appropriate for critical applications and widely varying load. What allowance is actually correct depends on the equipment, the warm-up duty, the operating conditions, the trap type and the manufacturer's guidance. The current edition of the ASHRAE Handbook, HVAC Systems and Equipment adds a table of steam trap sizing safety factors organised by application, which is the shape the question takes: an allowance belongs to a duty, not to steam in general.

Where warm-up governs, the practice is to calculate the warm-up load for the specific application and select on the larger of the two quantities rather than on the product. The alternative runs the other way: fix the warm-up time at which the warm-up load fits inside the capacity already selected, and put that time into the startup procedure as a controlled warm-up.

Per Spirax Sarco and TLV guidance on trap selection: the allowance applied to the running load stands in for a separately calculated warm-up load, and where warm-up governs, the selection is made on the larger of the two rather than on the product.

Differential Pressure Is Reported, Not Applied

The quantity the page puts the most emphasis on does not appear in the formula for the required capacity. That is correct rather than an oversight, provided it is clear where the differential does belong.

The required capacity is the running load times the factor, and it answers the question of how much condensate has to be removed. How much condensate forms is set by heat transfer. A coil condensing steam at a given duty produces the same mass of condensate whether the trap downstream sees 4 bar (58 psi) across it or 0.4 bar (5.8 psi), so the pressure across the trap has no influence on how much water arrives at it.

The differential decides the other side of the comparison. It sets how much condensate a particular trap will pass, and manufacturers publish capacity charts giving rated capacity as a function of it, with the same body passing substantially different quantities at different differentials. The differential is what takes the user to the correct line of the chart.

That has a direct consequence for the selected capacity field. What belongs in it is the rated figure at the differential that applies, not the largest figure in the table. Entering the largest inflates the ratio and can move the result into a favourable band the operating condition does not support, which produces an answer that is arithmetically correct and physically meaningless.

The order of operations is four steps, and the calculation performs three of them:

1. Calculate the required capacity from the load and the allowance.
2. Establish the differential pressure at the operating condition under consideration.
3. Read the candidate trap's rated capacity from the manufacturer's chart at that differential.
4. Compare the two.

Step 3 is the one that stays with the user, and it is the step that carries the manufacturer's data.

Which differential to use is the next question. Manufacturers state that the trap has to handle the load at the lowest differential relevant to the operating condition being considered. For a main warmed up while the line pressure is still rising, that lowest differential occurs during startup, which is also when the load is highest, and the two adverse conditions coincide. For equipment brought online against a main already at pressure they do not coincide, and treating them as though they always do overstates the requirement. The check is at the lowest differential belonging to the startup condition of the application in hand.

Per Armstrong and TLV selection guidance: required capacity follows from the condensate load while rated capacity follows from the differential pressure, so the pressure is what takes the user to the correct line of the manufacturer's chart rather than a term in the requirement.

Where the Rated Capacity Comes From

The number entered as the selected capacity comes off a chart, and the shape of that chart differs by trap type. The differences are not small enough to ignore.

Four things set what a trap passes. The orifice area available when the valve is open. The differential pressure across it. The state of the condensate, specifically how far it has been subcooled below saturation temperature. And the type of trap, which decides whether the valve is open continuously or only for part of the time.

A float trap opens its valve in proportion to the rate at which condensate arrives and discharges continuously, so its rated figure corresponds to a steady condition. An inverted bucket trap works in cycles, and its rated figure is an average across the cycle. A thermodynamic trap snaps open and closed, and its capacity falls away with back pressure more sharply than the others. A thermostatic trap does not discharge until the condensate has cooled below saturation temperature, so its behaviour during startup, when condensate arrives at saturation temperature, is not the behaviour its steady rating describes.

Two consequences follow. A figure taken from the chart for one type does not transfer to another type of the same connection size, because the size of the connection is not what sets the capacity. And the ratio the calculation produces belongs to whichever rated figure was entered, inheriting every condition attached to it, including the trap type, the differential and the assumed condensate temperature.

There are also things the chart does not contain. It does not describe behaviour at part load, which matters for cyclic traps because part load changes the cycle rather than the flow. It does not describe how the capacity changes as the seat wears. And it does not account for the orifice fouling, which is a normal service condition on systems carrying scale or pipe debris rather than an unusual one.

Per Armstrong capacity chart guidance and manufacturer data: rated capacity depends on valve orifice, differential pressure, condensate subcooling and trap type, so a figure taken from one chart does not transfer to another trap of the same connection size.

Back Pressure Eats the Differential

The downstream side of a trap is rarely at atmosphere, and everything that raises it reduces the pressure available to push condensate through.

Back pressure has four common contributions. Static head, where the condensate is lifted after the trap, which costs about one bar per ten metres of lift (about 1 psi per 2.3 ft). Pressure in the collecting header, wherever the return is not vented to atmosphere. Friction loss along the return line. And the pressure developed by flash steam in a header serving several traps at once.

The arithmetic is worth carrying out for an ordinary arrangement. Steam at 7 bar absolute (101.5 psia), discharging into a header at 2 bar absolute (29.0 psia), with a 5 m (16.4 ft) lift after the trap:

Differential if discharged to atmosphere: 7.0 − 1.0 = 6.0 bar (87.0 psi)
Header back pressure:                               2.0 bar (29.0 psi)
Static head of 5 m (16.4 ft) of water:              0.5 bar ( 7.3 psi)
Actual differential:      7.0 − 2.0 − 0.5         = 4.5 bar (65.3 psi)

The differential has fallen by a quarter against the atmospheric discharge case, and the trap has not changed.

What that does to capacity is read from the chart at 4.5 bar (65.3 psi) rather than derived from the pressure ratio. Capacity does fall with differential, but how steeply depends on the trap type and on the chart, and scaling the atmospheric figure by the ratio of differentials is not a substitute for reading the correct line. The calculation reports the differential and leaves that reading to the user.

Trap types differ in how much back pressure they tolerate before the behaviour changes rather than merely reduces. Thermodynamic traps stop working once back pressure exceeds a stated fraction of inlet pressure, and the manufacturer states that fraction for the model in question. Other types continue to operate at high back pressure and lose capacity while doing so, which is a different failure and a less abrupt one.

Where back pressure is high, the responses are to select a type less sensitive to it, to separate the collecting headers by pressure so that traps discharging at different pressures do not share one, and to use pumping arrangements where the condensate cannot return under its own pressure at all.

Per Spirax Sarco guidance on condensate return and manufacturer data: back pressure from static lift, return header pressure and flash steam reduces the differential available across the trap, and the resulting capacity is read from the chart at the reduced differential rather than scaled from it.

Flash Steam Occupies the Return Line

Condensate leaving a trap is at the saturation temperature for the pressure upstream of it. Downstream of the trap that temperature is above saturation for the lower pressure there, so part of the condensate flashes back into steam.

At 7 bar absolute (101.5 psia) the condensate is at 165 °C (329 °F). Discharged into a header at 2 bar absolute (29.0 psia), where the saturation temperature is 120 °C (248 °F), the surplus enthalpy has to go somewhere, and it goes into evaporating part of the water.

Flash fraction = (h_f1 − h_f2) / h_fg2

h_f1    liquid enthalpy at the pressure upstream of the trap,
        kJ/kg (BTU/lb): 697 (300) at 7 bar absolute
h_f2    liquid enthalpy at the pressure downstream,
        kJ/kg (BTU/lb): 505 (217) at 2 bar absolute
h_fg2   latent heat at the pressure downstream,
        kJ/kg (BTU/lb): 2,202 (947) at 2 bar absolute

(697 − 505) / 2,202 = 0.087, about nine percent by mass

Nine percent by mass is a small number and a misleading one, because volume is what a pipe carries. At 2 bar absolute the specific volume of the vapour is 0.8857 m³/kg (14.19 ft³/lb) against 0.00106 m³/kg (0.0170 ft³/lb) for the liquid, a ratio of about 836. Nine percent of the mass therefore occupies the overwhelming majority of the volume, and a return line sized for the volume of liquid is running nearly full of vapour at many times the intended velocity.

That closes a loop back to the trap. Velocity in the header produces friction loss, friction loss raises the pressure in the header, and header pressure is back pressure on every trap discharging into it. An undersized return line reduces the differential across traps that were selected correctly against a differential the return was supposed to provide.

The responses are to size the return on the volume of the mixture rather than the volume of the liquid, to separate the flash steam in a flash vessel and put it to use at the lower pressure, and to keep headers at different pressures separate.

Per Spirax Sarco guidance on flash steam and condensate line sizing: condensate discharged to a lower pressure flashes in proportion to the enthalpy difference, and the resulting vapour occupies most of the volume in the return line even at a modest mass fraction.

Oversizing Behaves Differently by Trap Type

Excess capacity is not uniformly harmless, and what it does depends on how the trap opens and closes.

A float trap modulates. The valve opens as far as the arriving condensate requires and no further, so excess capacity means the valve spends its life at small openings. That is tolerated reasonably well. Working at a small opening against a high differential does increase wear on the seat, because the whole pressure drop is taken across a small gap, but the discharge remains continuous and the trap remains in control.

Cyclic traps behave differently. A thermodynamic trap and an inverted bucket trap both work by filling, opening, discharging and closing. Excess capacity shortens the cycle, because the volume that has to accumulate before the trap opens is reached sooner relative to what the trap can pass. The number of operations per hour rises. Wear in these types is counted in operations rather than in hours, so an oversized selection consumes the service life faster than a correctly sized one at the same duty.

Thermostatic traps respond to condensate temperature and discharge once it has cooled below saturation. Excess capacity changes their picture less than it changes a cyclic trap's, but the accumulation of cooled condensate upstream of the trap is a property of the type and is present whatever the size. Whether that accumulation is acceptable is an application question, not a sizing question.

For the calculation's own bands, this means the OVERSIZED band names a condition whose consequences it cannot describe, because it never asked which type of trap is being considered. The same ratio of 1.8 means a float trap running at part opening and a thermodynamic trap cycling faster than it needs to, and those are not the same finding.

There is also a straightforward commercial side. A larger trap costs more to buy, and when it fails open it passes more steam than a smaller one would. Life cycle cost assessment weighs both against the purchase price.

Per TLV guidance on trap selection and life cycle cost: the consequence of excess capacity depends on whether the trap discharges continuously or in cycles, since cyclic traps wear by the number of operations and an oversized selection shortens the cycle.

The Failure Modes Point in Opposite Directions

Both ways of getting a trap wrong produce symptoms, and both sets of symptoms are easy to attribute to the wrong cause.

Insufficient capacity backs condensate up. It accumulates ahead of the trap and rises into the steam space. Heat transfer surface floods, so the duty delivered falls below the duty designed. Outlet temperature from the equipment becomes unsteady, because the flooded fraction varies with load. Where a length of main fills with water and steam then arrives at velocity behind it, the result is water hammer, which is a structural event rather than a performance one.

Excess capacity, or a trap that has failed open, passes live steam. The return line runs hotter than it should. Pressure in the header rises, which raises back pressure on every other trap discharging into it, so one failed trap degrades the traps around it. None of this announces itself. The plant keeps working, and the loss continues until somebody measures it.

The two are readily confused. Flooded heat transfer surface presents as insufficient capacity in the heat exchanger, which invites a conclusion about surface area or fouling. Steam passing through traps presents as higher fuel consumption, which invites a conclusion about heat loss or leaks. Neither conclusion points at the trap.

They are separated by measurement rather than by inference. Temperature across the trap is the first test, because in normal operation the downstream side is substantially cooler and when steam is blowing through the difference is small. Ultrasonic testing is the second, because a continuous signal indicates continuous flow where a cyclic trap should be silent between discharges. A sight glass, where one is fitted, settles it directly.

What follows for the calculation is a boundary. The ratio it produces describes the condition at the time of selection. What a trap is doing after some years in service is established by surveying it, and the two questions do not answer each other.

Per TLV and Spirax Sarco guidance on trap surveys: undersized traps flood the steam space while failed-open traps pass live steam, and the two are distinguished by temperature measurement across the trap and by ultrasonic testing rather than by the sizing calculation.

What the Ratio Does Not Test

A dimensionless ratio between two numbers the user supplied tests the arithmetic between them, and nothing about either of them.

What it establishes is exactly this: that the entered rated capacity exceeds the product of the entered load and the entered allowance, and by how much.

What it does not establish is a longer list. Whether the running load is right. Whether the allowance is enough for the startup condition of this application. Whether the rated figure was read at the differential that actually applies. Whether the trap type suits the service. Whether the type selected tolerates the back pressure it will see.

Every one of those five can be wrong while the ratio lands comfortably inside a favourable band. The ratio inherits the reliability of what was entered and adds nothing to it, which is true of every comparison of this shape and worth stating plainly because a green band reads like a verdict.

What the ratio is good for is comparison and screening. It compares candidate traps against each other on identical input assumptions, which is a real use. It catches order-of-magnitude errors, the kind where a capacity in pounds per hour has been entered against a load in pounds per minute. And it gives a reason to go back to the manufacturer's data when a candidate lands close to a band edge, since a band edge is where the difference between two candidates stops being decisive.

A favourable category should be read as a statement that the arithmetic is consistent, not as a statement that the selection is suitable. The page describes its bands as illustrative preliminary ranges, and that description is the whole of their standing.

Per the calculator's stated scope: the capacity ratio compares two figures the user entered, so it inherits their reliability and establishes nothing about the load, the allowance, the differential pressure basis or the trap type.

Worked Example: 1,200 Pounds per Hour at a Factor of Two

The imperial case matches the example carried on the calculator page.

Inputs. Running condensate load 1,200 lb/h (544 kg/h). Startup factor 2.0. Selected trap capacity 3,000 lb/h (1,361 kg/h), taken from the manufacturer's chart at the differential pressure that applies.

Step 1. Required capacity.

1,200 × 2.0 = 2,400 lb/h (1,089 kg/h)

Step 2. Capacity ratio.

3,000 / 2,400 = 1.25

Step 3. Category. A ratio of 1.25 falls between 1.15 and 1.50, so the band is RECOMMENDED. The band is an illustrative preliminary range on this page and not a rating from any manufacturer.

Step 4. What the allowance provides. The difference between the required and running figures is 1,200 lb/h (544 kg/h). Any startup load above that difference is not covered, and the two ways out of that are a larger allowance or a longer warm-up.

Step 5. What a load of this size corresponds to. At a latent heat of about 2,066 kJ/kg (888 BTU/lb), a running load of 544 kg/h corresponds to a thermal duty near 312 kW (1,065,000 BTU/h). That is the scale of a heat exchanger rather than a length of pipework.

Step 6. What the startup check involves here. For a heat exchanger the startup load is the sum of heating the metal and heating the product held in it, and for a liquid product the second term usually dominates. Both are calculated for the specific unit. The expression given earlier covers the metal only.

Step 7. What changes if the allowance changes. At a factor of 1.5 the requirement is 1,800 lb/h (816 kg/h) and the ratio becomes 1.67, which is OVERSIZED. At a factor of 3.0 the requirement is 3,600 lb/h (1,633 kg/h) and the ratio becomes 0.83, which is UNDERSIZED. The same trap at the same load visits three of the four bands on changes to one assumption, and that assumption is the one quantity here that is not measured.

Step 8. What the rated figure belongs to. The 3,000 lb/h belongs to the differential pressure it was read at. At a different differential the same trap model has a different capacity, and the ratio moves with it.

Step 9. What the result does not establish. That the trap type suits the service. That it tolerates the back pressure it will see. That a factor of 2.0 covers the startup condition of this particular unit.

Step 10. What to do with the result. Calculate the startup load for the equipment in question and select on the larger of that and the required capacity. Confirm the rated capacity at the lowest differential belonging to the operating condition being considered. Confirm that the type chosen is suitable at the expected back pressure.

Metric Example and the Warm-Up Check

The metric case also matches the page, and it lands in a different band for an instructive reason.

Inputs. Running load 900 kg/h (1,984 lb/h). Startup factor 1.5. Selected capacity 1,400 kg/h (3,086 lb/h).

Required: 900 × 1.5      = 1,350 kg/h (2,976 lb/h)
Ratio:    1,400 / 1,350  = 1.037, shown as 1.04
Category: 1.00 up to 1.15 → ACCEPTABLE

The selected trap exceeds the requirement by 3.7 percent. The page notes for this band that the margin available for startup and for varying load is limited, which is a fair description of three and a half percent.

The warm-up check is where this example becomes interesting. A running load of 900 kg/h at a latent heat of 2,066 kJ/kg corresponds to about 516 kW (1,761,000 BTU/h). For an insulated line losing of the order of 110 W/m (114 BTU/(h·ft)), that would need roughly 4,700 m (15,400 ft) of pipe on one trap, which no one builds. A load of this size belongs to a heat exchanger, and the startup load for a heat exchanger is calculated from the mass of the unit and the mass of the product inside it.

Apply the same factor of 1.5 to the pipework case from earlier and the picture inverts. The running load there is 19.2 kg/h (42 lb/h), so a factor of 1.5 gives a requirement of 28.8 kg/h (63 lb/h). The warm-up rate falls to that value at a warm-up time of about 116 minutes. A 30 minute warm-up of the same line needs 111 kg/h (244 lb/h), which is 5.8 times the running load.

One multiplier, two applications, and it is adequate in one and short by a factor of about four in the other. The quantity that decides is the mass being heated relative to the running duty, and nothing in the calculation asks for it.

Per Spirax Sarco guidance on warm-up load: the ratio of startup to running load depends on the mass being heated relative to the running duty, so an allowance appropriate to a heat exchanger does not transfer to a distribution main.

Application Boundaries: Selection, Startup, Verification

The model is a running load multiplied by an assumed allowance, compared against a rated capacity the user entered. Everything below sits outside it.

Startup load. Calculated separately from the mass of equipment being heated, the mass of product in it, the temperature rise, the latent heat and the warm-up time. The multiplier stands in for this quantity rather than computing it, and for the pipework case above the gap runs from 1.9 to 17.3 times the running load depending on the time allowed.

Differential pressure. Reported by the page and used to select the correct line of the manufacturer's capacity chart. It is not a term in the required capacity and does not need to be.

Trap type. Decides behaviour when oversized, at part load, and under back pressure. Float, inverted bucket, thermodynamic and thermostatic types differ in all three and the calculation asks for none of them.

Back pressure. Made up of static lift at about one bar per ten metres (about 1 psi per 2.3 ft), header pressure, friction in the return line and flash steam. In the worked arrangement above it took a 6.0 bar (87.0 psi) differential down to 4.5 bar (65.3 psi).

Flash steam. About nine percent by mass between 7 and 2 bar absolute (101.5 and 29.0 psia), occupying most of the volume in the return line and raising back pressure on every trap sharing the header.

Condensate subcooling. Affects capacity and differs by trap type, thermostatic types most of all, since they discharge only below saturation temperature.

In-service condition. Wear and fouling move actual capacity away from rated capacity. Establishing the present condition of an installed trap is a survey question, answered by temperature measurement across the trap and by ultrasonic testing.

The classification bands. Illustrative preliminary ranges belonging to this page, not manufacturer ratings and not an industry rule.

Final selection. Made from manufacturer capacity data for the specific trap type at the differential pressure that applies, with the warm-up condition checked separately.

Per TLV, Armstrong and Spirax Sarco selection guidance: multiplying a running load by an allowance and comparing it against an entered rated capacity is the scope of this model, while warm-up load, differential pressure basis, trap type, back pressure, flash steam and in-service condition each require separate treatment.

Steam Trap Capacity Calculator

Standards and References

• Spirax Sarco Steam Engineering Tutorials (Spirax Sarco, current online edition). Warm-up load and the steam consumption of pipework, steam trap selection by application, flash steam and the sizing of condensate return lines. This is the source for the treatment of warm-up load in this article.
• TLV, Steam Trap Selection guidance (TLV, current published edition). Safety factor in trap selection, the role of differential pressure and of back pressure, life cycle cost assessment and the consequences of excess capacity by trap type.
• Armstrong International, trap capacity and selection guidance (Armstrong International, current published editions). Published capacity charts by trap type and differential pressure, and the requirement that a trap handle the load at the lowest differential relevant to the operating condition.
• ISO 6704:1982, Automatic Steam Traps, Classification (International Organization for Standardization, 1982, reviewed and confirmed 2024). The classification of automatic steam traps by operating principle. The document covers classification and is not a capacity sizing method.
• ASME International Steam Tables for Industrial Use, Third Edition (American Society of Mechanical Engineers, 2014), based on IAPWS-IF97 (International Association for the Properties of Water and Steam, 1997; current revision R7-97(2012)). Saturation temperature, latent heat and liquid enthalpy at the pressures used in the warm-up and flash calculations here.
• ASHRAE Handbook, HVAC Systems and Equipment (American Society of Heating, Refrigerating and Air-Conditioning Engineers, 2024), Chapter 11, Steam Systems. Steam and condensate system arrangement and components. The current edition adds a table of steam trap sizing safety factors organised by application.
• ASME B36.10-2022, Welded and Seamless Wrought Steel Pipe (American Society of Mechanical Engineers, 2022, revising B36.10M-2018). Outside diameters and wall thicknesses of standard steel pipe, from which the 16.1 kg/m (10.8 lb/ft) used for DN100 standard wall in the warm-up calculation follows.
• ASTM C680-19, Standard Practice for Estimate of the Heat Gain or Loss and the Surface Temperatures of Insulated Flat, Cylindrical, and Spherical Systems by Use of Computer Programs (ASTM International, 2019). The calculation route for heat loss from insulated pipework. The 110 W/m (114 BTU/(h·ft)) used here is an order of magnitude for a line of this size and temperature difference, not a value taken from this document.
• Steam trap manufacturer data (current published editions). Capacity charts by type and differential pressure, stated back pressure limits for individual models, and behaviour at part load, none of which is derivable from a sizing ratio.

FAQ

Related Calculators

• Flash Tank Sizing: separating the flash steam from the condensate, which takes volume out of the return line and back pressure off every trap discharging into it.
• Condensate Pump Sizing: moving condensate where back pressure or geometry means it cannot return under its own pressure.
• Boiler Feed Pump Sizing: returning the recovered condensate to the boiler, which closes the steam and condensate loop the trap sits in.
• Condensate Return Line Sizing: sizing the return for the volume of the flashing mixture rather than the liquid, since the line that is too small is what raises the back pressure.
• Boiler Efficiency Calculator: the efficiency of the source, which sets what the steam lost through a failed-open trap actually cost to raise.
• Heat Exchanger Calculator: the equipment whose condensate the trap removes, and whose surface floods when the removal falls behind.
• HVAC Heat Load Calculator: the heat load from which the steam demand follows, and from it the condensate that has to be trapped.
• Duct Insulation Loss Calculator: heat loss through insulation, the same mechanism that sets the running condensate load of a steam main.