Mist Eliminator

What a Mist Eliminator Is and What Is Being Decided

The problem being solved is entrainment, liquid droplets light enough to ride the gas. Wherever gas and liquid meet at velocity, boiling, spraying, bubbling, condensing, or desuperheating, some fraction of the liquid ends up airborne, and gravity alone separates it only slowly and only for the larger drops. Downstream equipment pays for what stays airborne: condensers foul, vacuum pumps corrode, compressor blades erode, catalysts poison, stacks rain reagent, and product walks out of the vent. The mist eliminator is the engineered surface that intercepts this load inside the vessel the process already has, a pad, pack, or candle placed across the gas path so that every stream line passes within capture distance of a wire, a vane wall, or a fiber.

What is actually being decided in this purchase is four guarantees. First, separation performance: the element must meet the carryover target at the service flow, expressed either as removal efficiency at droplet size bands or as an exit loading in mg per Nm3 or ppmw, and the honest datasheet states the basis rather than a bare 99.9 percent. Second, hydraulic fit: the vessel diameter fixes the available area, and the element must deliver the performance at the design velocity that area implies, with a turndown story for the loads below design; mesh pads flood above their maximum velocity and lose capture below their minimum, so the window matters. Third, pressure drop: the element adds permanent resistance in series with the vessel, and the budget, tens to a couple hundred millibar, must be reconciled with the process, especially under vacuum services where every millibar costs steam. Fourth, survival in service: fouling, corrosion, and mechanical support decide whether the element is an asset in year five or a collapsed plug causing a shutdown, and the wash provisions, materials, and support design are purchased with the element, not improvised later.

Against its neighbors the mist eliminator occupies deliberate ground. A steam separator is an in-line pipeline device drying flowing steam by cyclonic or baffle action, bought as a vessel with nozzles; a mist eliminator is the internal installed inside process vessels and columns, and the two meet when a separator's internals are a mesh pad, but the products are purchased differently, internals by size and material to fit a known shell, separators as standalone vessels. A coalescing filter is a cartridge in a housing that removes aerosol from compressed gas by depth filtration and is replaced when loaded; the mist eliminator drains continuously and is not a consumable. A cyclone separator spins the whole flow to throw droplets and solids outward, effective for bulk load at moderate pressure drop, and often precedes a mist eliminator that polishes what the cyclone lets through. An oil water separator or an industrial filter or bag filter addresses liquid-liquid or dust duty and never competes with droplet capture in gas. The mist eliminator takes the narrow territory of continuous, self-draining droplet capture inside a vessel, and no neighbor holds that position.

Standards context is largely manufacturer engineering practice rather than a single component code, because the internals are non-pressure parts inside vessels that carry the code stamp themselves. Vessel fabrication around the element follows ASME Section VIII or EN 13445 as the shell requires, and the element supplier works to the vessel's nozzle schedule and support ring dimensions. Performance references come from manufacturer bulletins, such as the Koch-Glitsch mist elimination and phase separation literature, which publish the velocity constants, thickness, and efficiency data behind the sizing. Process services bring their own codes as context, steam systems their purity expectations, scrubbers their emission permits, and tank vents their environmental rules read against flame arrester and pressure vacuum relief valve requirements where the vessel is atmospheric storage; an inerted tank feeds the pad's underside through a tank blanketing valve, and any closed shell carrying the element protects itself with a safety relief valve sized without credit for the element's open area. A datasheet that quotes efficiency with no droplet basis, or velocity with no turndown, is a price tag, not a specification.

The RFQ should also name the vessel context so the element fits the first time. The shell diameter and orientation, vertical upflow or horizontal crossflow, fix the element geometry and the support and drainage arrangement; the nozzle schedule and the manway size decide how the element ships, in sections, coils, or whole, and a pad that cannot pass the manway is a site fabrication problem discovered at the worst moment. Services with solids or polymerizing mist need wash sprays above and below the element, supplied from a clean header with isolation, and the drain, a channel or trough returning to the liquid space below the gas inlet, must be sized so captured liquid actually leaves rather than re-entraining. Instrumentation around the vessel, a differential pressure transmitter across the element where fouling is expected, a level transmitter on the drum, a pressure transmitter on the gas space where the velocity basis and vacuum degree matter, and a flow meter on the gas, turns the performance into monitored numbers. Stating the whole context in the RFQ lets one supplier check velocity, support, and drain compatibility in a single pass.

How Mesh, Vane, and Fiber Elements Capture Droplets

Mesh pads capture by inertial impaction on wires. The pad is knitted wire, commonly 0.1 to 0.28 mm wire, looped into a sock and crimped into layers, giving a structure of perhaps 97 to 99 percent voidage: gas flows through nearly unobstructed, but every droplet has mass, and where the gas bends around a wire the droplet, reluctant to bend, crosses the streamline and hits the wire. Captured liquid spreads on the wire, joins films, grows to drops heavy enough to drain down the wires against the gas flow, and leaves the pad's bottom face. Because capture relies on droplet inertia, efficiency rises with droplet size and gas velocity and falls with both small; the classic mesh pad removes essentially all droplets above about 5 micron at design velocity, the majority above 3, and little below 1, where droplets behave nearly as gas. Velocity also has a ceiling: above it, the draining films re-entrain from the pad's downstream face and performance collapses, and below a minimum the inertia that drives capture fades; the sizing window, expressed through the Souders-Brown velocity constant k, is therefore a band, not a point, and it is the heart of the datasheet.

Vane packs capture by turning the flow instead of filling it with wires. Corrugated plates, typically spaced 15 to 40 mm, form zigzag or curved channels a defined distance deep, 150 to 300 mm family; gas accelerates around each bend, droplets leave the streamline and deposit on the walls, and hooks or drainage channels in the corrugations route the captured film to a pocket shielded from the gas so it drains without re-entraining. The vane pack tolerates velocities well above mesh limits, handles high liquid loads and liquid slugs, resists fouling far better because its passages are millimeters not wire gaps, and pays for it with a coarser capture, effective above roughly 8 to 10 micron, and slightly higher pressure drop at like conditions. It is the choice where the gas is dirty, the loads heavy, or the velocity window of mesh cannot be honored by the existing shell.

Fiber beds capture the mist that mesh and vanes cannot, the submicron aerosol from chemical processes, sulfuric acid plants, chlorine drying, and compressor lube oil carryover. Gas flows at very low velocity, centimeters per second family, through a deep bed, 50 to 150 mm, of very fine fibers, glass or polymer 5 to 30 micron; submicron droplets are captured by interception, where they touch a fiber in passing, and by Brownian diffusion, where their random walk walks them into a fiber regardless of the flow. Efficiency above 99 percent at 1 micron and below is achievable, at pressure drops from 100 to 400 millibar and climbing as the bed loads, and the beds drain the captured liquid slowly or, in the common candle form, hold it until blown down. The purchase logic is therefore staged: bulk capture by gravity or cyclone, coarse mist by mesh or vane, fine mist by fiber, each stage doing the fraction the previous cannot, and a specification that sends submicron mist at a mesh pad is asking the mesh for physics it does not own.

The sizing arithmetic deserves a worked look. The Souders-Brown relation gives the maximum superficial velocity as a constant k times the square root of the density difference ratio, and the constants are the supplier's product data: mesh pads in typical process services run k around 0.076 to 0.11 m/s family by style and service, vanes higher at 0.12 to 0.18, and the allowable velocity fixes the minimum vessel diameter at the design flow and density. A vertical drum passing 10,000 Nm3 per hour of gas at 1.2 bar and 40 deg C, gas density about 1.4 kg per cubic meter, needs about 2.6 square meters of free area at a mesh k of 0.08, a diameter near 1.8 meters; the same duty at half the density, hot gas near atmospheric, doubles the area and the diameter swells past 2.5 meters. Turndown matters as much: at 30 percent of design the mesh velocity falls below its effective minimum and fine capture decays, while a vane pack simply gets quieter. The RFQ handles both ends by demanding performance at design, maximum, and minimum stated loads.

Pressure drop and liquid load complete the mechanism picture. A clean mesh pad at design velocity drops about 25 to 75 millibar, vanes a bit more by depth, fiber beds as stated above; the numbers rise with loading, with fouling, and with liquid flood, and the differential pressure transmitter across the element is the trend that separates an aging element from a failing one. Liquid load, the volume of entrained liquid per volume of gas, is an input to sizing: pads absorb loads to roughly 10 to 20 cubic meters of liquid per hour per square meter family before flooding, vanes tolerate several times that, and above those numbers the vessel needs upstream bulk separation, a feed device or cyclone, before the polishing element is asked to work. The water balance logic is the same as in any separation train: each stage takes what the previous leaves, and the element is sized on the load it will actually see.

Types, Construction, and Duty Boundaries

Three element families cover the catalogs, in ascending order of fineness and cost. The knitted mesh pad is the general purpose workhorse: thicknesses of 100 to 150 mm standard and up to 300 for special styles, wire materials from 304 and 316 stainless through monel, hastelloy, and titanium for aggressive service, and plastics, polypropylene and PTFE, where corrosion or non-stick surfaces dominate; styles vary wire diameter and weave density to trade pressure drop against capture, and composite pads layer a coarse style below with fine above. The vane pack serves dirty and high load duty: stainless or plastic plates in galvanized or alloy frames, horizontal or vertical flow orientations, with hook and pocket drainage geometries by product family. The fiber bed candle or panel serves submicron mist: glass or polymer fiber in supporting cages, candles of 200 to 300 mm diameter and 1 to 3 meters long hung in banks, or flat panels in grids, with dedicated drain and blowdown connections.

Construction details carry the warranty across all three families. Mesh pads ship as coils or sections assembled in place with the vessel's support ring or grid below and a hold-down above; the support must carry the wet pad weight plus the differential pressure load at maximum flow, and a pad sucked through its grid at upset is the classic failure of an under-supported retrofit. Vane packs bolt into frames that transfer load to the shell, and the sealing between pack and wall, gasket or welded baffle, decides whether gas bypasses the element, because a 5 percent bypass sends the carryover target out the window regardless of element efficiency. Fiber candles hang from a tubesheet with individual or common drains, and the sealing of each candle to its seat is the field check that decides whether the bed ever sees the flow. In every family the shipping split, the manway the sections must pass, belongs on the drawing the supplier works to.

Boundaries are worth stating bluntly. A mist eliminator is not a dust collector: solids load blinds the fine families, and a bag filter or dry collector, not a wetted mesh, is the instrument for dry particulate, though mist eliminators downstream of wet scrubbers catch the slurry droplets that carry solids with them, which is the standard coexistence. It is not a separator vessel: it is bought to fit a vessel, with diameter, support, and nozzles defined by the shell it lives in, and a buyer who orders a pad without the vessel drawing owns the fit risk. It does not separate liquid-liquid, so free oil in water wants an oil water separator, and it does not dry steam in a pipeline the way an in-line steam separator does. And it cannot rescue a vessel with an undersized disengagement space: if the gas space between liquid surface and element is too short or the inlet velocity too high, the element receives a flood it was never sized to polish, and the correct fix is bulk separation upstream, not a thicker pad.

The duty map below consolidates the boundaries into one table a specifier can scan before writing the RFQ; it pairs common services with the element family that survives them and the choices that generate callbacks.

Key Specification Parameters

The table below decodes the datasheet lines that decide whether a quoted element matches the vessel and the service. Each row is a purchase defining parameter; the paragraphs after the table explain what to demand and what a weak answer looks like.

The efficiency row deserves a paragraph of its own because it is the row most often waved at. Ninety-nine percent is a meaningless number without the droplet basis, because a mesh pad is 99 percent at 5 micron and far less at 0.5, and a fiber bed is 99 percent at 0.5 and overkill at 5; the strong datasheet shows a grade efficiency curve or a table by size band, and the strong specification states the process's actual droplet distribution or its source, vendor estimate, plant test, or design assumption. Where a permit or guarantee reads exit loading, mg per Nm3, demand the conversion basis too, gas rate and density, so the number is auditable at stack test rather than arguable.

The velocity and liquid load rows decide whether the element fits the shell at all. Demand the design velocity with its k basis, the maximum velocity before flood, and the minimum at which capture holds, then check all three against the vessel's actual diameter and the flow range, minimum load included; a pad that floods at 110 percent of design in a vessel that sees 115 percent during upsets is a shutdown with a lead time. The comparison table below compresses the evidence checks into the form a specification meeting can argue from.

The support and wash rows close the specification. The support grid carries wet weight plus differential at maximum flow, and the drawing should show it; the drain arrangement, troughs, channels, or candle drains, must connect captured liquid back to the liquid space without crossing the gas inlet, and the nozzle and flange details for external drain connections belong with the element order where the vessel is new. In fouling service the wash header design, spray coverage, and isolation valves are the difference between a pad cleaned in place and a pad replaced at every outage. The quotation that quotes the element and leaves support, drainage, and wash as vessel exercises is not a complete quotation.

Selection Decision Factors and RFQ Language

The decision sequence is short enough to run as a checklist in a specification meeting, and each step eliminates a class of quotations before price comparisons begin.

1. Characterize the service first: gas composition and density at condition, flow at design minimum and maximum, pressure and temperature, and the liquid, its composition, loading, and droplet size knowledge or assumption.
2. Fix the carryover target with its basis, downstream equipment limit, permit, or product loss economics, in a number the supplier can be held to.
3. Choose the family on droplet physics: mesh for the 3 micron and above polishing, vane for dirty and high load, fiber for submicron mist, staged if the load spans bands.
4. Check the shell fit: available diameter against design velocity, gas space height, inlet arrangement, and the manway the sections must pass.
5. Budget the pressure drop against the process, with special care under vacuum where tens of millibar are real steam, and choose style accordingly.
6. Select materials against the actual chemistry, condensing or dry, chlorides, acids, or salt, with the exotic premium paid only where the corrosion case exists.
7. Scope support, drainage, and wash: grid or ring rated for load, drain returned below the inlet, sprays where fouling is credible, and DP monitoring where trend matters.
8. Fix the evidence set: grade efficiency, velocity window, exit carryover basis, clean and maximum pressure drop, materials certificates, and the section plan against the vessel drawing.

Against neighboring separation equipment, the comparison a specifier actually needs runs as follows, and it belongs here in the selection chapter rather than in the page title. Gravity space in the vessel itself removes the biggest drops free, and the first question is always whether more gas space solves the problem without an element; when it cannot, because the shell is fixed or the mist too fine, the element earns its place. A cyclone separator upstream removes bulk load cheaply and protects the element from floods. An in-line coalescing filter polishes compressed gas in cartridge form, a consumable answer for aerosol in dry gas, while the mist eliminator drains continuously in wet process gas. A shell and tube heat exchanger or plate heat exchanger downstream is the protected equipment in most evaporator stories, and the carryover target is usually written to protect one. The mist eliminator wins when continuous, self-draining capture inside an existing or new vessel is the requirement, and every alternative either consumables, bulk only, or new pressure parts prices higher. Choose on the droplet distribution, not on the family a supplier leads with.

The application map below turns that comparison into a table the project file can carry.

The RFQ block should be written once and reused. A complete block reads: service description with gas composition and density at operating condition; design, minimum, and maximum gas flow with units and basis; pressure, temperature, and vacuum degree where applicable; liquid composition and expected loading with its source, and droplet size distribution or the assumption to be used; required exit carryover in mg per Nm3 or ppmw with basis; element family and style preference or request for recommendation with justification; design velocity and k with flood maximum and capture minimum against the stated vessel diameter; thickness or depth and materials of wires, plates, or fibers and frames; support and hold-down arrangement against the vessel drawing with load rating; drain arrangement returning below the gas inlet; wash provision with header and isolation scope; pressure drop clean and maximum allowable; instrumentation provisions, differential pressure transmitter across the element; and the evidence set, grade efficiency data, section plan against manway, and material certificates EN 10204 3.1 where audited. A quotation returned without the efficiency basis, the velocity window, or the fit plan is incomplete at any price.

Installation, Standards, Failure Modes, and Maintenance

Installation quality decides whether the element performs as sized. Set the element level in vertical upflow vessels, because a tilted pad drains unevenly and floods its low side first, and seal the periphery against bypass, mesh with a gasket or stuffing strip, vanes with welded or gasketed baffle strips, candles at their tubesheet seats, since bypass defeats the whole purchase. Orient the drain correctly, troughs and channels discharging below the gas inlet or to a sealed return, never a open-ended drip into the gas space where it re-entrains immediately. Respect the gas space between the liquid surface or inlet and the element: the element polishes, and asking it to bulk separate a flooded space is the most common commissioning disappointment. Wash headers, where scoped, get installed with their isolation and are proven by coverage test at spray pressure before the vessel closes, because a wash header that misses half the pad is discovered at the first fouling event.

Commissioning practice proves the numbers. At design flow, record the differential pressure across the element against the clean datasheet figure, the baseline every later reading trends against; on vacuum services the reading resolution matters, and the transmitter range should be scaled to the element's budget, not the vessel's. Witness the carryover where a method exists, a target plate or catch pot on a probe, or rely on the downstream evidence, condensate cleanliness, compressor suction drum trends, stack opacity, and log it. Run the turndown check if the process allows, minimum load with the capture behavior observed through the same downstream evidence, and record the velocities against the supplier's stated window. The file that leaves commissioning is the baseline for every future decision about the element.

The standards context is vessel and process practice. The shell that carries the element is code stamped, ASME Section VIII or EN 13445, and the element supplier works to the vessel drawing and nozzle schedule; performance references are the manufacturer's published velocity constants and efficiency data, the Koch-Glitsch mist elimination bulletin family among them; services bring their own regimes, steam purity expectations on boiler and evaporator duty, emission permits on stack services, and the tank vent codes that pair with pressure vacuum relief valve and flame arrester hardware on atmospheric storage. A specification that states the vessel drawing revision, the service basis, and the evidence set by name transfers the burden of proof to the supplier, which is where it belongs.

Failure modes in service follow a short and predictable list. Fouling and blinding: scaling or polymerizing mist deposits in the element, differential pressure climbs, gas velocity through the remaining open area rises until flood and carryover appear together; the wash provision and the DP trend exist for exactly this, and the wash chemistry must dissolve the actual deposit, not just wet it. Corrosion: wires thin and break, the pad sags into its support and bypasses, or the whole element sheds fragments downstream; material selection against the condensing chemistry is the prevention, and a pad of the wrong alloy in acid condensate fails between outages, not across decades. Mechanical collapse: upset differential above the support rating, or liquid slugs at startup, crushes the element into the grid; supports sized for the upset case, not just design DP, are the countermeasure. Bypass: thermal distortion, shrinkage of plastic styles at temperature, or a poor retrofit seal opens a gap at the wall and carryover jumps while DP reads normal, the signature failure that sends investigation to the element when the gap is the culprit. Re-entrainment at overload: liquid load or velocity above window floods the pad and the outlet runs wet, fixed by upstream bulk separation or a family change, not by a thicker pad. Drain failure: a blocked trough or a seal lost from the drain line lets gas blow up the drain and strip the captured liquid straight through; the drain seal and its liquid leg are checked at every outage.

Maintenance intervals scale with service cleanliness more than calendar time. Clean services, steam and clean vapor, need an outage visual, support and seal check, and thickness sampling on the wires, yearly to turnaround cadence. Fouling services run the DP trend between outages and wash on schedule or on trend, with the wash chemistry reviewed against the deposit analysis each time it changes character. The DP baseline from commissioning plus the trend file converts every reading into a decision number: wash now, plan replacement, or investigate bypass. None of these checks costs a tenth of one unplanned shutdown from a collapsed or blinded element, and every one produces a number the next RFQ can use, which is how a purchasing file gets smarter instead of just older.