Demister Sizing: Why Western and Chinese Conventions Give Different Diameters

Why the Same Demister Duty Gives Different Diameters: Five Conventions That Move the Answer
11/15/2026
Why the Same Demister Duty Gives Different Diameters: Five Conventions That Move the Answer
11/15/2026

The same duty, two offices, two diameters: 0.77 m from one side, 0.52 m from the other. Neither engineer made an arithmetic mistake. One used a capacity factor of about 0.107 m/s and designed at 70 per cent of the resulting maximum velocity; the other used a coefficient of about 0.20 m/s and designed at 80 per cent. Both formulas are the same Souders-Brown form. The coefficients and the margin are not.

This page sets out where that gap comes from, how to convert between the two so that a quote can be compared, and what to write in a datasheet when the element will be fabricated in one country and specified in another.

The formula is shared; three things inside it are not

Every reference, Western and Chinese, writes the allowable velocity as a coefficient times the square root of the density ratio. What differs is what the coefficient means, which mesh form it was derived for, and how much of the maximum is taken as the design point.

Convention Coefficient, wire mesh Design fraction of maximum Pad height in practice
Western separator practice, standard knitted mesh about 0.107 m/s vertical, 0.128 horizontal 0.60 to 0.75 100 mm typical, 150 mm for higher efficiency
Western vane practice 0.152 vertical, 0.198 horizontal, up to 0.305 for pocketed 0.60 to 0.75 pack depth set by turns and spacing
Chinese design calculation for common mesh forms about 0.198 to 0.233 depending on mesh form 0.20 to 1.0 of maximum, commonly 0.80 150 mm preferred, 100 mm only where the efficiency demand is lower

The Chinese figures above are the coefficients published in widely circulated design calculations for knitted mesh forms, where the coefficient is tied explicitly to mesh construction, surface tension, viscosity and the droplet size to be captured. They are not the same coefficient as the Western separator value, because they were not derived on the same reference: mesh form, open area and pad thickness all move it, and a denser, thicker pad is a different hydraulic machine from a 100 mm pad of standard knit.

The regulatory anchors differ too. In China the wire mesh demister is specified through a dedicated standard covering mesh type, gas velocity and nominal diameter, with pad height, mounting style and single-stage pressure drop following from the same document; vane separators are covered separately, with a face velocity limit, a blade pitch range and a pressure drop expectation. In Western practice the velocity is usually left to the separator chapter of a data book plus the supplier’s own test data, with the vessel standard controlling structure rather than separation performance. That difference explains why a Chinese specification tends to state a velocity and a pressure drop, while a Western specification tends to state a duty and leave the velocity to the vendor.

What the gap looks like on one duty

Illustrative case: 1 m3/s of vapour, liquid density 1,000 kg/m3, vapour density 1.2 kg/m3, so the density term is 28.85.

Basis Maximum velocity Design velocity Area Diameter
Western mesh, 0.107 m/s, 70% factor 3.09 m/s 2.16 m/s 0.46 m2 0.77 m
Western vane, 0.152 m/s, 70% factor 4.39 m/s 3.07 m/s 0.33 m2 0.64 m
Chinese mesh, coefficient 0.201, 80% factor 5.80 m/s 4.64 m/s 0.22 m2 0.52 m
Chinese mesh, coefficient 0.233, 80% factor 6.72 m/s 5.38 m/s 0.19 m2 0.49 m

The two mesh rows differ by a factor of 1.47 in diameter, or 2.15 in area, and both are defensible engineering. Note what that means practically: a vessel sized on the Western basis will accept a pad made anywhere, whereas a vessel sized on the denser-pad basis is committed to that construction. Putting a 100 mm standard pad into the second vessel and expecting the same performance is the mechanism behind a large share of “the demister does not work” complaints on imported equipment.

Four things that move the number besides the coefficient

  1. Pad height and mesh form. A 150 mm pad of denser weave carries a higher coefficient than a 100 mm pad of standard knit, because the drainage path is longer relative to the gas path and the open area differs. A coefficient quoted without the pad construction is not a coefficient.
  2. Reference area. Vessel cross-section, element face area, or net open area. The peripheral gap between a round element and a round shell is typically 20 to 50 mm on each side, which on a 1 m vessel is 4 to 10 per cent of the area.
  3. Where the margin sits. A design fraction of 0.8 leaves much less room for turndown and fouling than 0.6. On the same coefficient that alone is a factor of 1.15 in diameter.
  4. Pressure drop as the hidden constraint. A thicker, denser pad carries a higher coefficient and a higher clean pressure drop. A pad specified to one pressure drop expectation and sized to the other practice’s velocity will miss both.

How to compare two quotes in ten minutes

Convert everything to one basis before discussing any other number:

  1. Express the design velocity at the element face in m/s: not a coefficient, not a fraction, not ft/s.
  2. Ask for the pad construction: mesh form, wire diameter, density, height. Then the coefficient can be checked against the supplier’s own test data rather than argued.
  3. Ask which area the velocity refers to, and deduct the peripheral gap and support ring.
  4. Ask for the clean and expected dirty pressure drop at the design velocity. A quote without a pressure drop has no drainage story.
  5. Ask what margin was applied and for which conditions: turndown, fouling, foaming, vacuum.

With those five, the 0.77 m and 0.52 m quotes become a discussion about pad construction and margin rather than a disagreement about arithmetic.

What to write in a datasheet for a cross-border purchase

  • Design velocity at the element face, in m/s, with the density pair used. This is the one number that cannot be misinterpreted.
  • Pad construction or an equivalent statement: mesh form, height, wire diameter, target clean pressure drop. If the supplier’s standard construction differs, the velocity is what must be re-derived, not the pad quietly downgraded.
  • Reference area and gap assumption, so the fabricated element and the sized vessel agree.
  • Acceptance test: pressure drop at design flow, visual inspection of drainage, and documentation of the mesh form actually used.

The point of all four is that a demister is bought as a hydraulic element, not as a piece of metal. Two practices that disagree by 47 per cent in diameter are not in conflict; they are describing different pads.

On the numbers used here

The coefficients for Western separator practice are the standard load-velocity values used throughout this site. The Chinese coefficients quoted are those published in circulated design calculations for specific mesh forms, presented as an illustration of the difference in basis rather than as the text of a standard: the normative values for mesh type, gas velocity and nominal diameter must be read from the current edition of the relevant Chinese standard, from which the pad height, mounting arrangement and pressure-drop expectation also follow. Any specification should cite the edition it was written against.

Frequently Asked Questions

Why do Chinese and Western demister designs give different vessel diameters?

Because the coefficient and the design fraction differ even though the formula is the same. Western separator practice uses about 0.107 m/s for standard knitted mesh and designs at 60 to 75 per cent of the resulting maximum; widely circulated Chinese design calculations use coefficients around 0.20 to 0.23 for specific mesh forms and design at up to 80 per cent. On the same duty that is a factor of about 1.47 in diameter.

Is the higher Chinese coefficient wrong?

No, it describes a different pad. The coefficient is tied to mesh form, open area and pad thickness, and Chinese practice typically specifies a 150 mm pad where Western practice often uses 100 mm. The coefficient cannot be transplanted without the construction that produced it, in either direction.

What happens if a 100 mm standard pad is installed in a vessel sized on the denser-pad basis?

The pad sees a velocity well above its own allowable value, floods, and re-entrains what it captured. Pressure drop rises and carryover increases at the same time. This is the mechanism behind most complaints about imported demisters that fail to perform: the vessel was sized for one pad and the pad supplied is another.

Which number makes two quotes comparable?

The design velocity at the element face in metres per second, together with the vapour and liquid densities used to derive it. A capacity factor alone is ambiguous in three ways at once: unit system, reference area, and whether the design margin is already inside it.

What pressure drop should be expected from a single-stage mesh demister?

Clean, at design flow, the order of 100 to 150 Pa for a single stage in normal service, with dirty or fouled operation higher and a rise over time being the early warning of blockage. A supplier who cannot state both the clean and the expected dirty pressure drop has not described the drainage behaviour of the element.

Leave a Reply