Vacuum Systems for EPP and EPS Moulding: Pump Type, Sizing and Leak Control

Quick answer: A bead foam moulding line only needs rough vacuum — typically somewhere in the region of 0.2–0.5 bar absolute — but it needs it fast, repeatedly, and while the pump is swallowing hot water vapour. That combination is what makes vacuum sizing different from ordinary industrial vacuum. The right system is a pump whose ultimate vacuum is comfortably deeper than the process target, a receiver tank that covers the few seconds of peak demand, a knock-out pot and pre-condenser ahead of the pump so it is not pumping live steam, and pipework large enough that the capacity you bought is the capacity that arrives at the steam chamber. Get those four right and vacuum stops being a variable in your cycle.

What vacuum actually does in a bead foam cycle

Vacuum appears several times in one EPP or EPS shot, and each appearance asks for something slightly different. Specifying a system without listing them is how plants end up with a pump that is right for one phase and wrong for the rest.

Note that two of those five phases move mostly water vapour, not air. That is the fact which decides the pump type and the size of everything in front of it.

Choosing the pump: three technologies that matter here

Vacuum generation for foam moulding almost always lands on one of three choices. Each is a compromise between ultimate vacuum, tolerance of water, and capital and running cost.

Liquid ring (water ring) pumps

The default choice on foam lines, and for a good reason: a water ring pump uses a rotating impeller to form a sealing liquid ring, and it is almost indifferent to swallowing condensate, foam residue and hot vapour. Its limit is physical rather than mechanical — the deepest vacuum it can reach is set by the vapour pressure of the sealing liquid. Water at roughly 15 °C has a vapour pressure of about 1.7 kPa (17 mbar); at 30 °C it is about 4.2 kPa (42 mbar). In practice that puts a well-maintained water ring pump in the region of 30–60 mbar absolute depending on seal-water temperature, which is comfortably deeper than the process target.

The consequence is worth underlining: a water ring pump that loses vacuum in summer is usually a seal-water temperature problem, not a worn pump. Recirculated seal water warms up, and the ultimate vacuum follows it up.

Water ejectors and steam ejectors

Ejectors have no moving parts and are cheap, and a water ejector is tolerant of condensate by definition. They are suited to modest vacuum levels and to plants that already have water pressure and drainage sorted. Their cost is water consumption and the effluent that comes with it, and their vacuum level is tied to the motive water conditions.

Dry running positive-displacement pumps (screw, claw, rotary vane)

These reach far deeper vacuum and do not need sealing liquid, which makes them attractive where the process genuinely needs it or where water discharge is a problem. They are considerably more expensive, they are far less forgiving of ingested condensate and bead fines, and they need filtration and knock-out that is properly engineered rather than nominal. On a foam line they are usually justified by a specific requirement, not chosen because "deeper is better".

Sizing: volume, target pressure and time

Vacuum capacity is a rate, and it is sized from three numbers you can get from the tool drawing and the machine specification: the chamber volume to be evacuated, the pressure you need to reach, and the time you have to get there.

For rough vacuum, where pump speed can be treated as roughly constant across the range, the first-order relationship is:

Worked example: a tool with a total chamber volume of 600 litres, evacuated from atmospheric pressure (about 1000 mbar absolute) to a 300 mbar target in 8 seconds.

That number is a floor, not a specification, for four reasons that all push in the same direction:

• Water vapour load. Condensate flashing to vapour inside a hot chamber can dwarf the air load. Pumps are rated on air; vapour has to be condensed or pumped as extra volume.
• Air released from the beads and the chamber walls. Beads carry air, and a wet chamber desorbs.
• Leakage. Every coupling, valve stem and gland leaks a little; a fleet of tools leaks more.
• Line resistance. See the pipework section below — this alone can remove 20–40% of the pump's useful capacity.

In practice, plants add a margin over the calculated figure and then verify it on the tool, because the vapour term is genuinely hard to calculate in advance. The honest way to express it is: calculate the air load, add margin for vapour and leakage, then confirm the evacuation time on the machine. That confirmation belongs in the same trial documentation as any other validated parameter in a T1 trial report.

One pump or one per press?

Central systems are cheaper per unit of capacity and easier to maintain, but they create a shared failure mode and a shared leak path — one badly connected tool drags down every press on the header. Dedicated pumps per press cost more and isolate failures. A common middle path is a central pump set with a generous receiver, plus isolation valves per press so a leaking tool can be valved off rather than tolerated.

When sizing a central system, do not assume every press peaks simultaneously, but do not assume they never will either. Machine controls that stagger the vacuum phase make a shared system work well; uncontrolled staggering is how an adequately sized central pump still produces slow evacuation at peak.

Receiver tanks: buy storage separately from capacity

The peak vacuum demand in a foam cycle lasts a few seconds. Sizing a pump to meet the instantaneous peak means buying a large pump that idles for most of the cycle. A vacuum receiver — a buffer vessel held under vacuum — covers the peak and lets the pump run steadily at a lower average rate.

The sizing logic is straightforward: given a peak draw of ΔV litres over the peak window, and an acceptable pressure rise of ΔP during that window, the receiver volume follows from the gas law in its simplest form, V_receiver ≈ ΔV × P_system ÷ ΔP. A larger tank permits a smaller pump; a smaller tank forces a bigger one. The tank is cheap per litre of stored vacuum and the pump is not.

Two practical points: locate the receiver close to the demand so it does the buffering rather than the pipe doing it, and put a drain at the bottom, because it will collect condensate.

Condensate, knock-out and seal water: what actually kills pumps

If there is one section of this article to act on, it is this one. A foam-line vacuum pump spends its life downstream of steam, and what comes at it is hot, wet and slightly abrasive.

Knock-out pot and pre-condenser

A knock-out vessel immediately ahead of the pump removes liquid carry-over and gives entrained droplets somewhere to fall out. A pre-condenser (usually water-cooled) goes further: it condenses vapour before the pump has to move it, which reduces the pumped volume dramatically and is often the cheapest capacity increase available on an existing line. Both need drainage, and both need to be sized, not guessed.

Seal water quality and temperature

For water ring pumps, seal water is a process variable. Three things matter: temperature (sets the ultimate vacuum), flow (sets the ring), and quality. Hard water scales the impeller and the casing; recirculated water warms and carries fines. A cooler or a heat exchanger on the seal-water loop, plus filtration and a degree of water treatment, is routinely the difference between a pump that holds vacuum year-round and one that is "fine in winter".

Carry-over of bead fines

Bead fines and any release agent residue that gets drawn through the vents end up in the pump. A strainer ahead of the pump is cheap insurance, and cleaning it should be on the same schedule as mould maintenance.

Pipework: the capacity you specify is not the capacity you get

This is where most vacuum specifications quietly fail. A pump's nameplate speed is measured at the pump inlet. What matters is the effective speed at the chamber, and the two are related by the resistance of everything in between:

Practical rules that follow from that:

• Oversize the vacuum line relative to instinct. What looks generous on a water line is often restrictive for gas at a few hundred mbar.
• Keep runs short and direct. Every metre costs capacity.
• Do not create low points. Condensate collects in dips and forms a liquid slug that chokes the line intermittently — the classic "vacuum is fine, then it isn't" complaint.
• Slope the line back to a drain or provide drain legs at low points.
• Count the valves. A partly closed isolation valve, a restrictive check valve or a clogged strainer can be the whole problem.
• Remember the vents. The tool's vent area is part of the path. Air and vapour have to get out of the cavity before the pump can remove them, which is the mould-side half of the same problem described in the cycle time levers.

Leakage and measurement

Vacuum leaks on a moulding line are not dramatic — they do not hiss like compressed air — and they are paid for slowly, in longer evacuation, wetter parts and a pump that never seems big enough. They are also easy to find.

Measure absolutely, not relatively

Gauge vacuum expressed as −0.7 bar reads differently with a change in atmospheric pressure and differently again with a tired gauge. Absolute pressure removes both arguments: specify the process target in mbar or kPa absolute, and put calibrated absolute gauges where decisions are made. This is also what makes a set point written in one plant reproducible in another.

Leak-rate test

Isolate the vessel or the line, pump it down to the process target, close the isolation valve and watch the pressure rise over a known period. The rise rate, multiplied by the volume, gives a leak rate in mbar·L/s. Do it per tool, not only on the header — a single tool with a perished coupling can mask itself in a central system's margin.

Finding the leak

Work from the most likely to the least: utility couplings and quick connects, valve stems and glands, sensor ports and drain points, then flanges. Soap solution still works for rough vacuum, ultrasonic leak detection works on gas leaks that are inaudible, and a temporary blank on a suspect branch settles arguments quickly. Fix the coupling standardisation problem at the same time — the keyed, position-fixed connections described in the mould-to-machine interface prevent most of this class of leak from recurring.

Commissioning and handover checks

What to put in the specification or RFQ

Vacuum is usually excluded from tooling quotations and then discovered as a plant cost. If you are buying a line or adding presses, the following should be explicit so that quotes are comparable:

• Chamber volume per tool and the evacuation target in absolute pressure.
• Required evacuation time, stated as a process requirement rather than a pump rating.
• Assumed vapour load, or at least an agreed margin over the calculated air load.
• Pump technology, with seal-water supply temperature and quality specified where relevant.
• Receiver volume and vacuum-relief provision.
• Knock-out and pre-condenser scope, with drainage.
• Line bore and routing limits, including a prohibition on low points.
• Isolation per press, so one tool can be valved off.
• Measurement: absolute-pressure gauges, test points and calibration.
• Performance acceptance: pump performance measured to a recognised method such as ISO 21360-1, witnessed at site conditions.

Most of those lines cost nothing to write and a great deal to add later. They belong in the same conversation as what drives mould cost — not because they change the tool price, but because they change whether the tool runs at the rate the quote assumed.

Frequently asked questions

Related reading: where EPP and EPS cycle time is really spent, how to fix common moulding defects, quick mould change and take-out automation, how to read a T1 trial report, what drives EPP and EPS mould cost and mould maintenance, cleaning and storage. See our EPP/EPS shape moulding machines, steam chambers for EPS moulds, EPS filling guns and customised EPS/EPP moulds, or review the key steps of designing an EPP moulding tool.