Steam Supply for EPP and EPS Moulding Plants: Boiler Sizing and Steam Quality

Q: Why is my boiler delivering less steam than its rating?

Because boiler ratings are quoted from and at 100 degrees C, assuming 2,257 kJ per kilogram of steam at a feedwater temperature of 100 degrees C. Real conditions need more heat per kilogram: at 8 bar gauge with feedwater at 20 degrees C, each kilogram needs about 2,688 kJ, so actual output is roughly 84% of the rating. Heating feedwater to about 90 degrees C lifts that to roughly 94%, and a deaerator at about 105 degrees C to roughly 97%. Specify the boiler on actual evaporation at your working pressure and feedwater temperature, not on the F and A figure alone.

Q: How much flash steam comes off returning condensate?

The flash fraction is the sensible heat at the higher pressure minus the sensible heat at the lower pressure, divided by the latent heat at the lower pressure. Condensate from 9 bar absolute flashing to atmosphere works out at roughly (743 - 417) / 2,257, or about 14%. That steam should be captured in a flash vessel and used on the lowest-pressure load available - space heating, hot water or the deaerator - rather than vented.

2026-10-26Industry News32 min read

Quick answer

How to size a boiler for EPP and EPS moulding, why steam dryness fraction changes fusion, and how to trap, recover condensate and treat feedwater.

Quick answer: A bead foam plant's steam system is judged on three numbers, and none of them is the boiler's nameplate rating. They are: the steam flow the boiler must actually deliver at the peak of a moulding cycle, the dryness fraction of the steam that reaches the machine, and how much condensate comes back. A shell boiler sold as “1,500 kg/h” is rated from and at 100 °C, which at 8 bar on cold feedwater is closer to 1,260 kg/h; wet steam condenses badly and quietly lengthens every cycle; and condensate sent to drain carries away roughly a quarter of the energy you have just paid for. Size on measured peak flow, specify dryness at the machine rather than at the boiler house, and design the condensate return as part of the system instead of as an afterthought.

EPP and EPS shape moulding machine with its steam inlet, cooling water and condensate connections — Steam arrives at the machine through one connection. Everything upstream of that point — boiler, header, separator, traps, reducing valve, insulation — decides whether what arrives is steam or a steam-and-water mixture.

Key takeaways

Pressure is a proxy for temperature. Saturated steam at 8 bar gauge is 175 °C; at 6 bar it is 165 °C. Losing a bar in an undersized main is losing degrees at the mould face.
Boiler ratings are quoted on a fixed basis. “From and at 100 °C” assumes 2,257 kJ per kg. Real conditions need more heat per kilogram, so the usable output is lower — typically 84–97% of the nameplate, depending on feedwater temperature.
Size on peak flow, then check turndown. Moulding steam demand is sharply cyclic; a boiler chosen only on average flow will be too small at the peak and will short-cycle at the trough.
Dryness fraction is a process variable. At 0.95, five percent of what passes the meter is water: it carries no latent heat, it fills the chest slowly and it makes one part of the mould hotter than another.
Superheat is not an upgrade here. Saturated steam condenses at constant temperature with a very high heat transfer coefficient. Superheated steam must give up its superheat first and does not help a foam mould.
Steam reaches the part. In bead moulding the steam is injected into the cavity and passes through the bead mass, so boiler water chemistry is not confined to the boiler house.
Traps fail open far more often than they fail closed, and a trap passing live steam is invisible unless somebody is looking for it. A trap survey is the cheapest efficiency measure in the plant.
Condensate at 8 bar still carries about 743 kJ/kg — roughly a quarter of the enthalpy of the steam that produced it, plus the cost of the water treatment that made it.
A mass balance finds leaks before a bill does. Feedwater in should equal steam out plus blowdown plus condensate lost. When those numbers disagree, something is passing.

Why steam, and why the pressure gauge is really a thermometer

Bead moulding is a heat-transfer process disguised as a moulding process. The beads have to be raised to the temperature at which their surface skins fuse, held there long enough for the fusion to develop through the wall, and then cooled. Steam is used because it delivers that heat at a constant temperature: saturated steam condensing on a metal surface gives up its latent heat without changing temperature, so as long as pressure is held steady the mould face sees a fixed, predictable boundary condition. That is exactly what a process controlled by fusion needs.

The consequence is that you cannot read a steam system from its pressure gauge alone — you have to read the gauge as a thermometer. Saturated steam pressure and saturation temperature are locked together:

Saturated steam properties in the range a bead moulding plant normally works in. Temperature, not pressure, is what sets fusion
Pressure (bar g)	Absolute (bar)	Saturation temp.	Latent heat h_fg (kJ/kg)	Sensible heat h_f (kJ/kg)
3	4	144 °C	2,133	604
5	6	159 °C	2,086	670
6	7	165 °C	2,064	697
8	9	175 °C	2,030	743
10	11	184 °C	1,998	781
12	13	192 °C	1,970	815

Two practical readings come straight out of that table. First, in this band one bar of pressure is worth roughly five kelvin of mould temperature, which is why a main sized a size too small — or a strainer left to block — shows up as a slow cycle rather than as an alarm. Second, as pressure rises the latent heat available per kilogram falls, so high-pressure steam is not automatically better steam: you are buying temperature, and you are buying it at the cost of energy per kilogram plus tighter traps and heavier pipework.

What pressure you actually need is set by the bead and the part, and it should be confirmed with the machine builder rather than copied from a table. As a planning guide, most EPS shape-moulding machines run on saturated steam in the region of roughly 0.6–1.0 MPa (6–10 bar gauge), while EPP and ETPU — denser beads and a tougher fusion requirement — often need the upper half of that range or a little above it. Whatever the number is, the important thing is that it must exist at the machine inlet during the heating phase, not at the boiler house during a quiet moment.

Superheat is not an upgrade for a foam mould

Superheated steam is valuable in a turbine and a nuisance in a mould. A superheated gas must first shed its superheat before it can condense, and while it is doing that its heat transfer coefficient is far lower than that of condensing saturated steam. The result at a mould face is a slower, less even temperature rise and a tendency to dry and overheat the outer bead layer while the core of the wall is still cold. Unless a machine builder specifies otherwise, ask for saturated steam and spend the money on dryness and pressure stability instead.

Keep the boundary clean. This article stops at the machine's steam connection and covers the boiler house, the distribution network, trapping and condensate. What happens to that heat inside the tool — steam chest design, bidirectional and cross-steam sequences, how much of the cycle is heating versus cooling — belongs to where EPP and EPS cycle time is really spent. The plant cooling loop is a separate system, covered in cooling water systems for EPP and EPS moulding plants. Vacuum and compressed air are separate again, in vacuum system selection and sizing and compressed air quality and demand.

Step 1: size the boiler on peak steam flow, not on machine count

Moulding steam demand is not a steady draw. It arrives as a pulse at the start of every cycle, when steam is admitted to fill the chest and drive air out of the bead mass, and then falls away almost to nothing through the cooling phase. A plant that adds up machines and multiplies by an average will build a boiler that is too small for the only moment that matters.

There are two honest ways to get the number. On an existing line, read it: a steam meter on the boiler outlet, logged at short intervals, will show you both the peak and the shape of the pulse. On a new line, ask the machine builder for steam consumption per cycle in kilograms, and the duration of the steam phase — from those two figures the peak flow follows directly.

Peak flow per machine (kg/h) ≈ steam per cycle (kg) ÷ steam-phase duration (h)
Design peak = Σ (coincident machine peaks) + pre-expander + space heating + margin

Worked example. A six-machine cell. Each machine takes about 28 kg of steam during a 40-second steam phase, so its own peak is 28 ÷ (40/3,600) ≈ 2,520 kg/h — but only for 40 seconds. Over a 180-second cycle that is an average of about 560 kg/h per machine, and with the cooling phase idle it is lower still. Cycles are not synchronised, so in practice perhaps three machines overlap at any instant: a credible coincident peak is 3 × 2,520 ≈ 7,600 kg/h for a few seconds. That spike is what a steam accumulator or a generously sized boiler steam space is for. The sustained peak you must size for, over a minute or two, is the average draw of all six running machines plus the pre-expander:

Building a design steam load for a six-machine bead moulding plant. Figures are illustrative; substitute your own machine data
Demand item	Basis	Design contribution
Six shape-moulding machines	About 560 kg/h average per machine while running, allowing for the cooling phase	3,360 kg/h — not the instantaneous spike above
Coincidence allowance	Cycles drift out of step; assume a proportion of machines heat at once	Apply a diversity factor agreed with the machine builder rather than assuming all six
Pre-expander	Batch steam demand for expansion, intermittent	Add the full batch peak; it can be large and short
Space heating, hot water, deaerator	Seasonal, steady	Add the winter figure, not the summer one
Distribution losses	Uninsulated mains, passing traps	Allow a real percentage — then go and eliminate it
Margin	Fouling, a future machine, a colder winter	Add last, and record it as margin rather than hiding it in the load

Now convert that load into a boiler rating — carefully

Boilers are conventionally rated “from and at 100 °C” (often written F&A), which means the quoted evaporation assumes feedwater at 100 °C producing saturated steam at 100 °C, requiring 2,257 kJ per kilogram. Real plants do not run at that condition: feedwater arrives colder than 100 °C and steam leaves at a higher pressure, so each kilogram of real steam costs more heat than the rating basis allows for. The usable output is the rating multiplied by an evaporation factor:

Actual output = F&A rating × [2,257 ÷ (h_g − h_feedwater)]
At 8 bar gauge: h_g ≈ 2,772 kJ/kg, feedwater at 20 °C ≈ 84 kJ/kg → factor ≈ 0.84

What a boiler rated 1,000 kg/h from and at 100 °C actually delivers at 8 bar gauge, depending on how the feedwater is heated
Feedwater condition	Feedwater enthalpy	Evaporation factor	Real steam output
Cold make-up, about 20 °C	≈ 84 kJ/kg	≈ 0.84	840 kg/h
After an economiser, about 90 °C	≈ 377 kJ/kg	≈ 0.94	940 kg/h
After a deaerator, about 105 °C	≈ 440 kJ/kg	≈ 0.97	970 kg/h

This is the single most common sizing error on a new foam plant. A boiler is bought on its F&A number, the plant needs the F&A number, and from the first cold morning the boiler runs flat out and never catches up. Two fixes, and they are complementary: heat the feedwater (which is worth roughly 10–13% of fuel as well as capacity, and is usually the better investment), and specify the boiler on actual evaporation at your working pressure and feedwater temperature in the enquiry, not on F&A alone.

Then check turndown and response

A boiler that is correctly sized at peak may still behave badly at the trough. A shell boiler typically modulates over a limited range — roughly three or four to one is a common order of magnitude — below which it cycles on and off. Every start purges the furnace with air, and that purge is heat straight up the stack, so a boiler spending its day short-cycling is paying a real efficiency penalty for capacity it does not need at that moment. Where the load swings hard, the options are: two smaller boilers sequenced so one runs in its efficient band; a steam accumulator (an insulated pressure vessel holding hot water that flashes to steam when header pressure dips, absorbing the cycle peaks and refilling between them); or accepting the cycling and understanding its cost. What you should not do is size for the peak and assume the trough looks after itself.

Step 2: steam quality — dryness fraction is a moulding variable

Steam quality, or dryness fraction, is the mass of vapour divided by the total mass of the mixture. A dryness fraction of 0.95 means that for every kilogram passing the meter, 950 g is vapour and 50 g is liquid water travelling at the same speed. That water is not harmless: it carries only its sensible heat, it has to be boiled off inside your mould before it can do any useful work, and it arrives unevenly.

Useful latent heat per kg = x × h_fg
At 8 bar gauge with x = 0.95: 0.95 × 2,030 ≈ 1,929 kJ/kg, against 2,030 for dry saturated steam — a 5% loss before anything else goes wrong

The energy loss matters less than the distribution of that loss. Water thrown out of a high-velocity stream tends to land where flow is slowest, which in a steam chest means corners, the far side of the cavity and the shadowed side of inserts. The result is a mould face that is hot in some places and wet in others, and that produces exactly the defect pattern described in how to fix poor fusion, short fills and warpage: patchy fusion, a heavy side and a light side, and part weight that moves between cavities. It also loads every trap downstream of the problem and invites water hammer, which is how steam headers lose their supports.

Where water in the steam comes from, and what to do about each source
Source	Why it happens	What fixes it
Carryover from the boiler	Water level too high, total dissolved solids too high causing foaming, or steam space too small for the evaporation rate	Correct level control; control TDS by surface blowdown; a properly sized boiler steam space. On a badly foaming boiler, a mechanical steam separator after the boiler is the reliable answer
Condensation in the mains	Steam cools in an uninsulated or badly trapped header; the film of condensate is dragged along by the flow	Insulate, and trap at every low point, every riser bottom, before every reducing valve and at the end of each main
Pressure reduction without separation	Reducing steam pressure increases its volume and velocity; any water present is atomised and carried further	Fit a separator with its own trap immediately upstream of the reducing valve — not somewhere else in the line
Wet steam from the boiler outlet	A boiler run near its maximum evaporation rate produces wetter steam	Headroom on the boiler, a steam separator, or both. Do not run a boiler at 100% of rating continuously and expect dry steam
Traps passing or blocked	A passing trap wastes energy but a blocked trap floods the line behind it	Trap survey with an ultrasonic or thermal instrument, on a schedule, recorded

How to find out what you actually have. Dryness can be measured at the point of use with a throttling calorimeter, or by sampling through a separating calorimeter; both are established test methods and both can be done on a live plant by a steam services contractor. If a measurement is impractical, the practical proxy is behaviour: a main that needs to be drained constantly, chest pressure that climbs slowly at the start of every cycle, and water hammer on valve opening all point the same way. Write the target into your specification as dryness at the machine inlet — commonly 0.95 or better — because a dryness figure quoted at the boiler outlet tells you nothing about what survived the journey.

Step 3: distribution — pipe size, trapping and pressure at the machine

Steam mains are sized on velocity, not on pressure rating. Saturated steam expands enormously as pressure falls, so the same mass flow occupies far more volume at 6 bar than at 10 bar, and a main that was fine at the higher pressure becomes a restriction after a plant expansion.

Volume flow = mass flow × specific volume | Pipe area = volume flow ÷ velocity | D = √(4A/π)

Worked example. 1,100 kg/h of steam at 8 bar gauge, where the specific volume is about 0.215 m³/kg. That is 1,100 × 0.215 ÷ 3,600 ≈ 0.066 m³/s. At a design velocity of 25 m/s the required area is 0.066 ÷ 25 ≈ 0.0026 m², giving a bore of about 58 mm — so DN 65 is the arithmetic minimum and DN 80 is the sensible choice once you allow for wet steam and future load. Distribution mains are usually kept to roughly 15–30 m/s; stay at the lower end where the steam runs wet, because velocity is what carries water along with it.

Steam trap types and where each one belongs on a bead moulding plant
Trap type	How it works	Best used for	Watch out for
Float and thermostatic	A float opens a valve proportional to condensate level; a thermostatic element vents air	Mould chests and heat exchangers — anywhere condensate must not be held up, because it discharges continuously	Water hammer and freezing can damage the float; needs a strainer and a check valve
Thermodynamic	Flash steam re-closes the disc after each discharge	Mains drip points, tracer lines, high-pressure headers	Cycles rather than modulating, so not ideal where condensate must be removed continuously
Inverted bucket	A bucket sinks and opens the valve as condensate fills it; steam makes it float and close	General duty on saturated steam; robust against water hammer	Can lose its prime; poor at venting air at start-up
Balanced pressure thermostatic	A capsule set close to saturation temperature discharges condensate and vents air	Mains ends, air venting at start-up, small loads	Sub-cools condensate slightly, so not ideal where maximum heat release is wanted

Trapping discipline is the same everywhere and it is not optional: a drip point at the end of every main, at the bottom of every riser, upstream of every reducing or isolating valve, and at intervals along long horizontal runs — commonly every 30–50 m. Each one needs a strainer, a check valve where condensate is lifted, and a means of testing it without shutting the plant down.

Reducing valves belong at the machine

Carry the pressure you need for the highest-pressure user, then reduce locally rather than distributing at the lowest common pressure. There are two reasons. Distributing at low pressure means large pipes carrying high-volume, low-density steam; distributing at high pressure means a smaller, cheaper network. And because throttling saturated steam is close to isenthalpic, dropping pressure at the point of use actually leaves the steam slightly superheated, which improves dryness rather than hurting it — provided, again, that a separator and trap sit immediately upstream so that no free water is being atomised into the reduced-pressure line.

Then insulate everything. An uninsulated main at 175 °C is not merely an energy loss; it is a condensate generator, and every kilogram of condensate it makes has to be trapped, cooled and (if you are sensible) returned. Insulation on mains, flanges, valves and the boiler itself is among the fastest-payback items in the whole plant, and it protects people as well as heat.

Step 4: condensate recovery and flash steam

Condensate leaving a trap at 8 bar gauge is water at about 175 °C, carrying roughly 743 kJ/kg of sensible heat — about 27% of the enthalpy of the steam that produced it. Sending it to drain throws that away, and it throws away treated water at the same time. It also sends hot water into a drain that may not be rated for it.

Condensate cannot simply be pushed back into a boiler feed tank at atmospheric pressure without releasing steam, because water at 175 °C is far above the boiling point at atmospheric pressure. The fraction that flashes is easy to calculate:

Flash fraction = (h_f,high − h_f,low) ÷ h_fg,low
From 9 bar absolute (743 kJ/kg) to atmospheric (417 kJ/kg, h_fg 2,257): (743 − 417) ÷ 2,257 ≈ 14%

So roughly one kilogram in seven of that returning condensate arrives as flash steam. That steam is not a problem to be vented away — it is free heat. Pass the condensate through a flash vessel and use the flash steam on the lowest-pressure load available (space heating, a hot water set, or the deaerator), then return the remaining liquid. A plant that vents flash to atmosphere and then buys fuel to heat the same feedwater has given away about 14% of its condensate energy for nothing.

What a condensate return system needs, item by item, and why
Item	Purpose	Practical note
Flash vessel	Separates flash steam from liquid so each can be used at the right pressure	Size on the highest condensate pressure in the plant, not the average
Vented feed tank with a level control	Buffers between intermittent return and continuous boiler demand	Size for several minutes of peak evaporation; keep it hot and insulated
Feedwater pump set	Delivers to the boiler at above boiler pressure	Hot water pumps need NPSH checking at 100 °C or more — cavitation here is common and looks like a control fault
Conductivity or hardness monitor on the return	Detects contamination — a leaking heat exchanger or corroded mains putting salts into the feedwater	With a dump valve, so contaminated condensate can be rejected automatically rather than reaching the boiler
Return main sized on two-phase flow	Condensate and flash steam travel together, at far higher volume than liquid alone	The most commonly undersized line in the plant. Size the return for the flash, not for the water
Back-pressure conscious traps	Allow condensate to be lifted into a pressurised return	State the return pressure in the trap specification; a trap sized for atmospheric discharge will not lift

Return ratio is a target, not a virtue. Between 60% and 80% return is a realistic range for a well-run plant; 100% is not, because some condensate is genuinely lost to blowdown, venting and processes. Chasing the last few percent with complicated pumping arrangements usually costs more than it saves. But a plant returning 30% has a findable problem — usually blocked traps discharged to drain, or a return main too small to carry its own flash steam.

Step 5: feedwater and boiler water — and why it reaches the part

In bead moulding, steam is not confined to a jacket. It is admitted into the mould cavity and passes through the bead mass to fuse the surfaces, which means whatever the steam carries can end up on the moulded part. That single fact turns boiler water treatment from a boiler-house concern into a product-quality and, for food-contact or electronics packaging, a compliance concern. It is also why scale and carryover in this industry are not just efficiency issues.

Feedwater has to be brought to near-zero hardness and low dissolved solids before it enters the boiler. Softening removes hardness but leaves the sodium salts and the alkalinity; where make-up is high or the boiler pressure is high, reverse osmosis or demineralisation is used to remove the dissolved solids as well. Returned condensate is the best feedwater available — it is already pure and already hot — which is a second, quality-based argument for the recovery system above.

Boiler water parameters worth monitoring, and the reason each one is on the list. Actual limits come from the applicable boiler standard and your treatment programme
Parameter	Why it is controlled	What goes wrong if it is not
Feedwater hardness	Hardness salts precipitate as scale on the hottest surfaces	Scale in a shell boiler is a fuel penalty and, at worst, a tube failure; softener failure must be alarmed, not discovered
Dissolved oxygen	Oxygen attacks steel as pitting, worst in the feed line and the boiler	Removed mechanically in a deaerator, with a chemical oxygen scavenger as the residual safeguard
Total dissolved solids / conductivity	Concentration drives foaming and carryover — the main source of wet steam	Wet steam to the mould, erratic level control, and deposits on the part
Alkalinity	High alkalinity causes caustic embrittlement and foaming	Boiler metal damage and again carryover
pH	Kept alkaline enough to passivate steel, not so alkaline as to attack it	Corrosion on one side of the range, embrittlement on the other
Silica	Volatilises with steam and deposits where pressure falls	Deposits on valves, orifices and the mould surface rather than in the boiler
Phosphate / polymer residual	Conditions sludge so it can be removed by blowdown instead of baking on	Sludge accumulates, hot spots form, efficiency falls
Condensate conductivity	The early indicator of contamination	Contaminants reach the boiler disguised as clean feedwater

Blowdown is how those concentrations are held steady: surface (or continuous) blowdown from the water surface, where dissolved solids concentrate, controlled from conductivity; and bottom blowdown from the lowest point, to remove sludge. Both waste heat, which is why blowdown heat recovery — flashing it down and using the flash, or recovering it through a heat exchanger into the feedwater — is worth doing on any plant with meaningful blowdown.

Where the steam touches the product, choose the chemistry accordingly. If moulded parts will come into contact with food, or are packaging for it, the boiler water additives need to be selected for that service. In the United States, boiler water additives used in steam that contacts food are regulated under 21 CFR 173.310; comparable rules exist in other markets, and many buyers will ask for a statement. This is a conversation to have with the water treatment supplier at the specification stage, not after a customer questionnaire arrives.

Compliance and inspection

A steam boiler is pressure equipment and sits inside a regulated regime that varies by market — the ASME Boiler and Pressure Vessel Code (Section I for power boilers, Section IV for heating boilers) in North America, EN 12953 for shell boilers and EN 12952 for water-tube boilers in Europe, with the Pressure Equipment Directive 2014/68/EU governing its supply. In practice this means: correctly sized and regularly tested safety valves, tested low-water and high-pressure cut-outs, a documented inspection and insurance regime, and operators who know what to do when the low-water lockout trips. None of that is optional and none of it interacts well with production pressure, which is precisely why it has to be scheduled rather than improvised.

Symptoms, causes and the first thing to check

Troubleshooting the steam side of a bead moulding plant
Symptom	Likely cause	Check first (cheapest to most involved)
Chest pressure climbs slowly at the start of every cycle	Wet steam, or a restriction between the header and the machine	Open the trap bypass briefly and listen; then compare header pressure with machine inlet pressure with the valve open. A large gap means a restriction, not a boiler problem
One side of the part fuses better than the other	Uneven steam distribution, water collecting on one side, or blocked vents	Vent condition in the tool, then dryness at the inlet. Do not start by raising pressure — see the defects guide for the full sequence
Water hammer when a valve opens	Condensate lying in the main; blocked or missing traps	Drip points at the low points and the end of the main; trap survey. Fix before something breaks
Boiler runs continuously and never reaches setpoint	Boiler sized on F&A rating, or feedwater too cold, or a large passing trap	Mass balance first: feedwater meter against steam meter against condensate return. Then re-check the evaporation factor
Fuel consumption rising with steady production	Passing traps, failed insulation, scale on the boiler heating surface, or air/fuel settings drifted	Trap survey, then flue gas temperature and oxygen trim, then water-side inspection
Part weight drifting between cavities	Steam supply varying between machines or across the cycle	Log header pressure through the cycle, not once per shift; compare with the part-weight trend in in-process quality control
White or chalky deposit on moulded parts	Carryover of boiler water solids, seen because the steam contacts the part	TDS and level control; check the separator; review the treatment programme
Condensate return ratio falling over months	Traps failing, or the return main progressively blocked by corrosion products	Return conductivity and flow; trap survey; inspect a section of the return main

Energy: where the money actually is

Steam is usually the largest single energy cost in a bead moulding plant, and the losses are unglamorous. In rough order of size on a typical site: condensate and flash sent to drain, cold feedwater, passing traps and uninsulated surfaces, carryover forcing longer cycles, and short-cycling at low load. Almost all of them are measurement problems before they are capital problems — the mass balance tells you the size of the prize before you spend anything.

For a sense of scale: 1,100 kg/h of steam at 8 bar, on feedwater at 20 °C, needs roughly 2,690 kJ per kilogram, so about 2.96 MW of heat. At 90% boiler efficiency that is around 3.3 MW of fuel input — on natural gas at roughly 36 MJ/m³ lower heating value, in the order of 330 m³/h at peak. Raise the feedwater from 20 °C to 90 °C and the heat needed per kilogram falls by about 11%, which is the same as finding 11% more boiler capacity for the price of an economiser.

Eight steps to specifying or auditing a plant steam system

Measure the load, or get it per cycle. A logged steam meter, or kilograms per cycle and steam-phase duration from the machine builder. Never a sum of nameplates.
State the pressure at the machine inlet that the process needs, then design the network to deliver it there during the heating phase.
Convert to a real boiler rating using the evaporation factor for your feedwater temperature and working pressure. Write “actual evaporation” into the enquiry.
Check turndown and peak response. Decide between sequencing two boilers, adding an accumulator, or accepting cycling — and know which you chose.
Ask for dryness at the point of use, commonly 0.95 or better, and require a separator with its own trap wherever pressure is reduced.
Size the mains on velocity and the returns on two-phase flow. Insulate mains, flanges, valves and the boiler; provide drip points to the discipline above.
Design condensate recovery with the boiler, not after it: flash vessel, vented feed tank, correctly sized return main, back-pressure conscious traps, and a contamination monitor with a dump valve.
Start a trap survey and a mass balance on day one and trend both monthly. These two records will find almost every problem in this article before it shows up as a quality complaint.

One project at a time, again. A steam-system upgrade and a cycle-time reduction both raise output per hour, and running them together makes the gain impossible to attribute. Fix the steam supply first — it is upstream, it is measurable, and it sets the ceiling on what the tool can do — then optimise the cycle following the sequence in where EPP and EPS cycle time is really spent.

Frequently asked questions

What steam pressure does an EPP or EPS moulding machine need?

It is set by the bead and the part, so the authoritative answer comes from the machine builder. As a planning guide, EPS shape moulding commonly runs on saturated steam around 0.6 to 1.0 MPa (6 to 10 bar gauge), and EPP or ETPU, with denser beads and a tougher fusion requirement, often needs the upper part of that range or slightly above. The number that matters is the pressure present at the machine inlet during the heating phase, not the boiler house pressure at a quiet moment, because roughly one bar of pressure in this range is worth about five kelvin of mould temperature.

Why is my boiler delivering less steam than its rating?

Because boiler ratings are quoted “from and at 100 °C”, assuming 2,257 kJ per kilogram of steam at a feedwater temperature of 100 °C. Real conditions need more heat per kilogram: at 8 bar gauge with feedwater at 20 °C, each kilogram needs about 2,688 kJ, so actual output is roughly 84% of the rating. Heating feedwater to about 90 °C lifts that to roughly 94%, and a deaerator at about 105 °C to roughly 97%. Specify the boiler on actual evaporation at your working pressure and feedwater temperature, not on the F&A figure alone.

What is steam dryness fraction and why does it matter in moulding?

Dryness fraction is the mass of vapour divided by total mass, so 0.95 means 5% of what passes the meter is liquid water. Water carries no latent heat, so useful heat per kilogram falls by about 5%. More importantly, the water lands unevenly, concentrating in slow-flow regions and on the shadowed side of inserts, which produces patchy fusion, uneven part weight between cavities and water hammer in the headers. Specify dryness at the machine inlet, commonly 0.95 or better, because a figure quoted at the boiler outlet says nothing about what survived the distribution network.

Would superheated steam improve moulding?

No, and it is usually counterproductive. Saturated steam condensing on the mould surface gives up latent heat at constant temperature with a very high heat transfer coefficient, which is exactly the boundary condition a fusion-controlled process needs. Superheated steam must first shed its superheat before it can condense, and while doing so it transfers heat poorly and tends to dry and overheat the outer bead layer while the core of the wall is still cold. Unless a machine builder specifies otherwise, ask for saturated steam.

How much condensate should a plant return, and why bother?

Between 60% and 80% is a realistic range for a well-run plant; 100% is not achievable because some condensate is lost to blowdown and venting. The reason to bother is that condensate at 8 bar gauge still carries about 743 kJ/kg, roughly a quarter of the enthalpy of the steam that produced it, and it is already-treated water. A plant returning only 30% almost always has a findable fault: traps discharging to drain, or a return main too small to carry its own flash steam.

How much flash steam comes off returning condensate?

The flash fraction is (h_f at the higher pressure minus h_f at the lower pressure) divided by the latent heat at the lower pressure. Condensate from 9 bar absolute flashing to atmosphere works out at roughly (743 − 417) ÷ 2,257, or about 14%. That steam should be captured in a flash vessel and used on the lowest-pressure load available — space heating, hot water or the deaerator — rather than vented.

How do I know if a steam trap has failed?

By testing it, on a schedule, with an instrument. Traps fail open far more often than closed: a passing trap discharges live steam continuously, which is invisible from outside, never shows up as a fault, and appears only as fuel consumption. Ultrasonic and thermal trap testers are quick to use on a live plant, and a trap survey with a written record is usually the fastest-payback item in the steam system. A blocked trap is easier to notice — it floods the line behind it — but it does just as much damage to cycle time.

How do I find steam leaks and losses without a consultant?

With a mass balance. Feedwater into the boiler should equal steam out plus blowdown plus condensate lost, over any period long enough to average the cycling. If feedwater exceeds the sum of the other three, something is passing — a trap, a relief valve, a leaking flange or an unrecorded take-off. Meter the feedwater and the condensate return, take the steam total from the boiler's own instrumentation, and trend the gap. It is the cheapest diagnostic in the plant and it never gives a false positive.

Does boiler water treatment matter for the moulded part itself?

Yes, because in bead moulding steam is injected into the cavity and passes through the bead mass rather than staying in a jacket. Anything the steam carries — carryover solids from high TDS, treatment chemicals, corrosion products — can deposit on the part. That makes carryover a quality issue as well as an efficiency one, and it means the treatment programme should be selected with the end use in mind: where moulded parts contact food, boiler water additives used in steam that contacts food are regulated, in the United States under 21 CFR 173.310.

Do I need a steam accumulator?

It depends on how sharp the peaks are relative to the boiler's ability to respond. A steam accumulator is an insulated pressure vessel holding hot water that flashes to steam when header pressure dips, absorbing the cycle peaks and recharging between them. It is worth considering when a boiler sized on sustained load cannot cover the short coincident spikes, when the boiler would otherwise short-cycle at low load, or when several machines start cycles together. Where the peaks are modest or the cycles well staggered, a properly sized boiler steam space and a sensible steam main often do the same job for less money.

Cooling Water Systems for EPP and EPS Moulding Plants: Tower Sizing, Pumps and Water Treatment

2026-10-23

Compressed Air for EPP and EPS Moulding Lines: ISO 8573-1 Purity, Demand and Distribution

2026-10-22

In-Process Quality Control for EPP and EPS Moulding: SPC, Part Weight and Traceability

2026-10-21

Need a custom EPP / EPS mold for this application?

Send us your drawing, target density and annual volume — our engineers reply with a tooling proposal, lead time and quotation within 24 hours.

Get a quotation