Mist Eliminator Replacement Without Drawings

Replacement enquiries often involve incomplete drawings, an old pad of uncertain history or an as-built vessel that has changed. A replacement can be specified from physical evidence, but only if the evidence route, unresolved discrepancies and remaining uncertainty are controlled before manufacture.

“No drawings” is three different problems

Treat these separately, because they carry different risk:

1. No drawing exists. You have a vessel, an old pad and operating history. Everything must come from measurement. Highest measurement burden, but no conflicting information.
2. A drawing exists but is not accessible. Held by a licensor, EPC or OEM under confidentiality, or archived in a format nobody can open. You may still get a support-ring elevation or manway detail from the vessel drawing even when the internals sheet is missing.
3. A drawing conflicts with field evidence. Record the discrepancy rather than declaring either source correct. Check drawing revision, measurement method, vessel modification and deformation, then freeze the accepted as-built basis in writing before manufacture. The physical opening controls fit, while the controlled design basis still governs dimensions and interfaces that cannot be changed informally.

State which case you are in when you send an enquiry. It changes what has to be measured and what can be assumed.

Decide first: copy the old design, or fix it

Make this decision explicitly before treating the old pad as the manufacturing basis.

If the old pad reached an explained service endpoint and there is no unresolved drawing conflict or material process change, its construction can be a useful starting point. If it failed early or for an unknown reason, copying it may reproduce the failure. Classify the evidence before ordering:

• Plugging with rising pressure drop — density or style may be unsuitable for the actual solids and liquid load. Increasing density without checking fouling and drainage can worsen the problem. See demister fouling and clogging and mesh pad vs vane pack.
• Pitting, cracking or uniform thinning — points first to a material and environment investigation, while deposits, stress, temperature and galvanic contact still need checking. Geometry alone will not cure a chemistry-driven failure; see demister pad materials selection and demister corrosion upgrades.
• Sagging, torn sections or collapsed mesh — may involve support, hold-down, handling, vibration, liquid loading or operation above the intended hydraulic range.
• Carryover with a clean, undamaged pad — cleanliness alone neither proves nor excludes a pad problem. Check bypass, inlet distribution, liquid level, drainage, mesh construction and actual velocity before reordering.

Copying an unexplained failed design can repeat the part cost, outage work and lost operating time.

Evidence to collect

Each item below exists because it constrains the design. Collect what you can; incomplete evidence still beats a blind catalogue guess.

Vessel and support geometry

• Vessel inside diameter at the demister elevation, measured at two or three positions around the circumference. An older or modified vessel may be out of round, and a pad cut to a single nominal diameter can leave a bypass gap.
• Support ring inside and outside diameter, ring width and thickness, and whether it is continuous or segmented.
• Available vertical space between the support ring and the next obstruction above. This caps total pad thickness including grids.
• Hold-down arrangement: top grid, clamps, tie rods, or nothing at all.

The manway constraint

Manway clear opening diameter is not a detail — it determines segmentation. A pad that cannot pass the manway cannot be installed regardless of how correct the rest of the design is. Record the clear opening, not the nominal flange size, and note any davit, internal piping or tray above that limits the path. Section width must be set from access dimensions, safe manual handling, support arrangement and sealing strategy rather than an unsourced universal figure. Segmentation is covered further in demister installation types.

The old pad itself

• Overall thickness, and thickness of the mesh layer separately from the grids.
• Section count, shape and chord dimensions.
• Grid bar cross-section, spacing and material; tie-wire arrangement and pitch.
• Wire diameter, measured with a micrometer on several strands from a clean area.
• A clean, undamaged offcut if one can be released — a physical sample answers more questions than any photograph.

Material identification

Look for a nameplate, stamped tag, heat number or paint mark before assuming anything. Where nothing is legible, the alloy has to be treated as unverified until it is either tested or specified afresh. Positive material identification on the old part is optional and separately quoted; it can indicate measured alloy chemistry at the test location, but it does not establish every component, the original traceability chain or what the duty requires.

Photographs that are actually useful

Take the pad in place from the manway before removal, then on the ground from both the grid side and the mesh side. Photograph the support ring and wall interface, section joints, hold-down details, any label, and the fouling or corrosion pattern with a scale reference in frame. The demister inspection checklist covers this as an outage procedure.

Estimating density from a sample

Knitted mesh density is often missing from replacement records. A clean sample can provide an estimate, but the result must be interpreted with material and construction data.

The method is direct: take a clean sample of known dimensions, weigh it, and divide mass by the volume it occupies as installed. That gives an apparent density in kg/m³. HG/T 21618-1998 Table 4.0.2 gives 128–186 kg/m³ for its listed styles, calculated at a material density of 7,930 kg/m³; HG/T 21586-1998 Annex A gives 98–182 kg/m³ for its listed drawer-type constructions, including round wire of 0.19 mm and 0.23 mm. The combined 98–186 kg/m³ span is not a universal acceptance or anomaly band. Before comparing another alloy or a polymer, correct for material density and confirm that wire form and knitting construction are comparable. Treat any measured result as an estimate to be confirmed, for three reasons:

• Fouling inflates mass. Gypsum, salts, polymer or iron sulphide deposits can add a large fraction of the sample weight. Weigh a cleaned sample, and say whether it was cleaned.
• Compression changes volume. A pad that has been clamped, sagged or crushed no longer occupies its original thickness.
• Corrosion removes metal. A thinned wire gives a low density and a low wire diameter that reflect the end of service life, not the original specification.

Combine the density estimate with material identity, wire diameter, wire form and mesh appearance rather than relying on any single figure. HG/T 21618-1998 provides structural, dimensional and preliminary-selection guidance for wire mesh demisters and covers a nominal diameter range of DN300–DN5200; HG/T 21586-1998 is the related drawer-type standard. Those references make a quotation easier to compare, but they do not by themselves decide final sizing or performance — see mesh density and thickness under HG/T 21618.

OEM style designations: what you can and cannot ask for

Replacement enquiries frequently arrive quoting a style number such as a York-series designation. This needs care from both sides.

Those designations originate with an equipment manufacturer and are protected marks — DEMISTER® is a registered trademark of Koch-Glitsch, with heritage in the Otto York product line. A style number is a useful shorthand for a mesh construction, and it is legitimate for you to describe the construction you need. What it is not is a purchase of the OEM’s product.

So be explicit about which of these you want:

• The OEM part. Buy it from the OEM or its authorised channel. Nobody else can supply it as a genuine article.
• A non-OEM replacement made to agreed physical parameters. Specify wire diameter, apparent density, mesh layer thickness, crimp and layering, material, grid and support details, then evaluate conformity to that frozen parameter set. Physical conformity does not by itself prove equivalent process performance.

Meshium quotes non-OEM replacements on the second basis only. Style references identify the existing equipment or the requested construction; the frozen manufacturing basis is the parameter set written into the quotation and PI, not a brand name. This distinction matters more, not less, when the end user’s specification names a proprietary style.

Fit-up is where replacements fail

A pad with correct density and material can still perform poorly if vapour goes around it. On replacements based on field evidence, possible causes include a wall gap on an out-of-round shell, section joints that do not close, missing hold-downs and distorted grids. A controlled drawing and shop trial fit can reduce bypass risk, but without complete as-built data or site trial fit they cannot guarantee vessel fit before shipment. This is particularly important where site access is constrained, as discussed in offshore separator demister replacement.

Practical measures worth agreeing before fabrication: overlap or a sealing strip at the wall, a defined section joint detail rather than butt-cut edges, hold-down grids supplied with the pad instead of reused, and a marked assembly sequence so a crew working through a manway installs sections in the intended order.

Verification before it ships

For a replacement controlled from field evidence, the agreed sketch becomes the manufacturing reference document. What is included in the quoted supply price should be stated in the quotation and PI, and for replacement work normally comprises a dimensional inspection record taken against the PI-agreed sketch at defined measurement points and acceptance limits, the ordinary material or batch traceability record named in the PI, product and packing photographs before shipment, and a packing list.

Two things are worth separating clearly. Specifying the material — the alloy grade for metal wire and hardware, or the polymer type together with the raw material producer, commercial designation and temperature limit for monofilament products such as PP, PTFE or ETFE — happens in the RFQ and is frozen in the PI as the manufacturing basis. Proving it happens at shipment through the traceability record named in that PI. Enhanced documentation under EN 10204 Type 3.1 or Type 3.2, third-party inspection and additional testing are optional, must be raised before material procurement and production so that traceability, cost and lead time can be confirmed, and are quoted separately.

Where fit certainty matters more than schedule, a paid sample or single-section trial may be quoted before the bulk order. Any sample-cost credit must be stated in the sample PI. For a fixed turnaround, compare that cost and schedule with the remaining fit risk before choosing the route.

Common mistakes

• Reordering “the same as last time” without classifying why the last one came out.
• Measuring vessel diameter at one position on an out-of-round shell.
• Sending a nominal manway size instead of the clear opening.
• Deriving density from a fouled or crushed sample and treating the number as the original specification.
• Increasing mesh density to cure carryover that is actually caused by velocity, distribution or level control.
• Asking for a branded style number and assuming the supplier is offering the branded part.
• Leaving material documentation to be decided after production, when the required records or inspection participation may no longer be recoverable.

If all you have is a phone photo of a damaged pad, send it as preliminary evidence. It may help identify the next measurement or document needed, but it is not enough to freeze a manufacturing basis by itself.