Why Demister Pads Collapse or Blow Out: Mechanical Failure Guide

Mist Eliminator Material Selection: SS316L, Duplex, Hastelloy or PTFE?
11/09/2026
Vane Type Inlet Device: Function, Sizing and Selection Rules
11/09/2026
Mist Eliminator Material Selection: SS316L, Duplex, Hastelloy or PTFE?
11/09/2026
Vane Type Inlet Device: Function, Sizing and Selection Rules
11/09/2026

Short answer: a demister pad almost never fails because the mesh is the wrong grade. It fails because the gas velocity exceeds what the pad can hold, because the support structure lets the pad sag or lift, because the pad does not touch the vessel wall and gas bypasses it, or because liquid arrives in slugs instead of as a mist. Mesh fragments found downstream are the result of one of those four, not of corrosion alone.

The five failure modes and how to tell them apart

Before ordering a replacement pad, identify which failure actually happened. A pad that is replaced without fixing the cause fails again within the same turnaround interval, and the second failure is usually worse because the fragments damage downstream equipment.

Failure mode What you find on inspection Underlying cause Corrective action
Lift-off / blow-out Pad displaced upward or downstream, mesh intact but out of position Gas velocity above the design value, or support grid spacing too wide so the top layer lifts Reduce face velocity or increase area; add intermediate support grids at 100-150 mm spacing
Sagging / collapse Pad bowed downward, densified at the centre, high local pressure drop Support grid spacing exceeds what the pad can bridge, or liquid load exceeds drain capability Shorten grid spacing, add support beams, review liquid loading and drainage
Wall bypass Pad looks clean while carryover persists; gap or channel at the vessel wall Pad diameter undersized, missing wall ring, ovality of the vessel not allowed for Fit with a wall ring or oversized pad trimmed on site; verify pad-to-wall contact all round
Fragmentation Loose wire downstream, filter or blower fouling, mesh torn at the edge Corrosion thinning of the wire, vibration, or mechanical damage during installation Upgrade alloy, add restraint, inspect installation method
Plugging / fouling Hard scale, polymer or salt build-up, pressure drop rising while flow is steady Solids carryover, crystallisation, or polymerised process material Wash system, upstream solids removal, higher open-area grade

1. Gas velocity above the design value

The capacity of a mesh pad is set by the Souders-Brown relationship, where allowable velocity depends on the density difference between liquid and gas and on an empirical K value. For a knitted wire mesh demister in vertical gas flow, K typically sits between 0.03 and 0.06 m/s. If the plant debottlenecked, added a blower, or operates at lower pressure than the original design, the actual velocity rises and the pad starts to lift liquid back off the mesh.

The tell-tale is that the pad fails at high load and behaves normally at turndown. That pattern points at velocity, not at the mesh grade. The fix is either more area, a lower K value (denser pad, more pressure drop), or a different internal type such as a vane pack that tolerates higher velocity.

2. Support structure and grid spacing

A standard pad is 100 to 150 mm thick and is supported on a grid. When the grid spacing is too wide, the pad sags between supports, the local density increases, pressure drop rises in that zone, and liquid accumulates until the section floods. Widely spaced grids also allow the upper layers to lift under velocity, which then redistributes the whole bed.

Practical checks during inspection: measure the actual pad thickness against the drawing, look for a visible dip at the centre of each grid opening, and confirm the grid has not deflected. If the grid itself is bent, the load exceeded its design and the spacing needs to be reduced, not just the pad replaced.

3. Wall bypass

A pad that leaves a gap at the vessel wall is effectively a hole in the separator. Gas takes the path of least resistance around the edge, and carryover continues even though the pad looks clean and undamaged. This is common after a replacement where the pad was ordered to the nominal vessel diameter without allowance for shell ovality, internal attachments or a wall ring.

The reliable fix is a wall ring or an oversized pad trimmed on site so it contacts the shell around the full circumference. Confirm contact at four quadrants at minimum, and check that nozzle openings or manways do not create an unprotected gap.

4. Corrosion and material loss

Wire diameters of 0.15 to 0.28 mm lose strength quickly once corrosion starts. A pad that has been in wet chloride service with SS304 will thin at the wire level long before the pad visually fails. Thinning shows up as mesh that no longer holds its shape, edges that crumble when touched, and loose wire downstream.

Where the process includes chlorides, wet acid or high temperature, the material question is decided first and the mesh grade second. See the material comparison in Mist Eliminator Material Selection for the alloy ranges and temperature limits.

5. Liquid slugs and thermal cycling

Pads are designed for dispersed mist, not for slugs. A sudden carryover of bulk liquid, a foaming event upstream, or a washwater surge compresses the bed, and repeated cycles progressively densify it until it plugs. Thermal cycling adds a second mechanism: differential expansion between the pad and the support grid loosens the fit and opens bypass paths.

Where slugs are a known process condition, the internal selection should move to a vane pack with a drainage path, or the upstream vessel should be given a knock-out section. Replacing the pad with a denser pad makes slugging worse, not better.

Inspection checklist before replacement

  • Measure gas velocity at operating conditions against the design K value.
  • Record pad thickness at the centre and at the edge; a difference above 15% indicates sagging.
  • Check pad-to-wall contact around the full circumference.
  • Inspect support grid deflection and spacing against the drawing.
  • Test wire diameter at several points to detect corrosion thinning.
  • Check drainage passages for blockage and confirm the liquid outlet is not submerged.
  • Look downstream for wire fragments, which indicate fragmentation rather than bypass.

When the answer is not a new pad

If inspection shows velocity above design, wall bypass, or slugging, a like-for-like pad replacement will fail again. In those cases the correct scope is a change in internal type or a change in the support arrangement, and the replacement element is only part of the fix. Sizing the element properly is covered in How to Size a Demister for a Given Gas Flow Rate, and hydraulic limits in Demister Flooding and Re-entrainment.

Send the inspection findings, the operating pressure and temperature, the measured velocity and photographs of the failed element, and we will return a written assessment of the failure mode with the corresponding fix.

Frequently Asked Questions

Can a demister pad blow out even if it was sized correctly?

Yes. Sizing sets the velocity limit, but blow-out is often caused by support grid spacing, missing wall rings or a liquid slug rather than by velocity alone. A correctly sized pad on a wide grid can still lift.

What causes wire fragments to appear downstream?

Fragmentation rather than bypass. Wire thinning from corrosion, vibration at a loose edge, or mechanical damage during installation releases loose wire that travels downstream. Inspect the pad edge and test wire diameter to confirm.

Should I replace a collapsed pad with a denser one?

No. A denser pad increases pressure drop and liquid retention and makes collapse more likely. Fix the support spacing, area or liquid load first.

How do I check for wall bypass without opening the vessel?

You cannot reliably. Carryover that persists with a clean pad and normal pressure drop is the symptom, but confirmation requires an internal inspection of pad-to-wall contact.

Worked example: how far an upset has to go before the pad floods

Take the pad sized in the standard case: 1.2 m diameter, 150 mm bed, 150 kg/m³, air at 60 °C and 1 atm carrying water droplets, design flow 10,000 kg/h.

Step Calculation Result
Maximum allowable velocity 0.107 × √((983 − 1.06) / 1.06) 3.26 m/s
Design velocity 0.70 × 3.26 2.28 m/s
Velocity at 150% of design flow 1.5 × 2.32 3.48 m/s = 107% of u max
Velocity at 180% of design flow 1.8 × 2.32 4.17 m/s = 128% of u max

The margin is smaller than it looks. At 150% of the design flow the pad is already above its capacity factor, which means the liquid film on the mesh can no longer drain against the gas drag: liquid holdup rises, pressure drop climbs, and droplets are re-entrained from the pad itself. The outlet gets wetter at exactly the moment the plant is least able to tolerate it.

That is why the 60 to 75% design band exists, and why a demister sized at 95% of u max has no upset margin at all. Where the process has a credible upset case above 150% of normal flow — a relief scenario, a compressor trip, a parallel train stopping — the pad should be sized on that case, not on the normal one.

Worked example: mechanical margin during a wash cycle

Condition Pressure drop across the pad What it loads on the support
Normal operation 0.14 kPa 0.14 kN per m² of face area
After a year of deposits 0.29 kPa 0.29 kN per m²
Wash nozzle starting over a dry pad, water slug passing through 2 to 4 kPa transient 2 to 4 kN per m², plus the weight of the water held in the mesh

A support designed for the normal operating load, with no allowance for the wash transient or for the wet weight of the bed, is the usual mechanical weakness behind a collapsed or dished pad. The failure is not caused by the process exceeding its design; it is caused by the support being designed against the steady-state number only.

Two checks are worth running whenever a pad has deformed: whether the support rating covers the wash transient, and whether the retaining arrangement can hold the bed against the upward drag at the highest velocity the vessel can actually see. A pad that lifts at the edge opens a bypass path at the wall, and no amount of mesh density recovers the efficiency lost through that gap.

Operational troubleshooting

How do you tell a collapsed pad from a fouled one?

A fouled pad shows a gradual pressure drop rise at constant flow with the outlet quality degrading slowly. A collapsed pad shows the opposite: pressure drop falls while carryover rises sharply, because gas is now passing through a thinned or channelled bed. The pressure drop direction is the discriminator, and a sudden change points to a mechanical event rather than fouling.

What support design actually prevents collapse?

A support grid sized for the full wet weight of the pad plus the transient load from washing, with beams at a spacing that keeps the unsupported span short, and a pad held down so it cannot lift. Most collapses start at an unsupported span or at a hold-down that was omitted, not at the pad itself.

Leave a Reply