Proper Application of Mist Eliminators

FUNDAMENTAL CONSIDERATIONS

PROPER APPLICATION of mist eliminators is based on understanding how they work. Vane and mesh devices both employ the same mechanism—known as inertial impaction—and thus are subject to the same basic design rules. Fiber mist eliminators, however, capture submicron droplets (those smaller than one micron) by an entirely different phenomenon—known as Brownian motion—leading to very different behavior.

Inertial Capture in Vanes

As shown in Figure 9, vanes bend the path of mistladen gas into relatively tight curves. As the gas changes direction, inertia or momentum keeps mist droplets moving in straighter paths, and some strike adjacent vanes. There, they are held by surface forces and coalesce (merge) with other droplets, eventually trickling down. If the vane material is wettable, a surface film promotes coalescence and drainage. In the case of upward flow, coalesced liquid disengages from the bottom of the vanes as droplets large enough to fall through rising gas. In the case of horizontal flow (Figure 10), the liquid trickles down vanes to a drain below.

Figure 9: Capture of mist droplets in a vane array with vertical flow	Figure 10: Vane array in a mesh-type mist eliminator

Inertial Capture in Mesh

In a mesh-type mist eliminator (Figure 11), each strand acts as an obstruction around which gas must flow. Within a very short distance upstream of a filament, the gas turns aside sharply, but some mist droplets are unable to follow. They strike the filament, adhere, and coalesce to form droplets that are large enough to trickle down and fall away.
	Figure 11: Droplet capture in a mesh-type mist eliminator

Inertial Capture Efficiency

Based on the principle of inertial capture, it is easy to understand the behavior of a vane or mesh mist eliminator in terms of the efficiency with which it captures mist droplets. Consider a droplet encountering a mesh strand or a bend in a vane. (To help imagine the relative dimensions involved in the case of a mesh pad, see Figure 12.) The following factors determine whether the droplet strikes the surface or turns and flows around with the gas:

Figure 11: Droplet capture in a mesh-type mist eliminator

1. Droplet size: The larger the droplet, the greater its momentum and the straighter its path when surrounding gas flows around an obstacle. Consequently, as seen in Figure 13, the efficiency of a given mist eliminator varies steeply with droplet size (keeping the same velocity and liquid and gas composition). For the example mesh pad made of 0.011-inch wire, efficiency jumps from nearly zero for 2-micron droplets to nearly 100% for 20-micron droplets. In a real situation, droplet sizes will be distributed over a range from less than one micron to well over 100 microns. The distribution curve may be narrow or broad, peaking anywhere within that range.

2. Strand diameter or corrugation spacing: The smaller the diameter of a mesh strand (or the closer the spacing between the corrugations of a vane), the more abruptly oncoming gas turns aside, and the more difficult it is for mist droplets to follow the gas. Thus, finer strands can capture smaller droplets (again assuming the same velocity and liquid and gas composition). This effect can be seen by comparing the three curves in Figure 13, representing mesh pads having different strand thicknesses. The 279-micron (0.011-inch) wire is 90% efficient for 6-micron droplets, compared to 3-micron droplets for the 152-micron (0.006-inch) wire and 1.5- micron droplets for 10-micron co-knit glass fibers.

Figure 13: Examples of variation of droplet capture efficiency with droplet size and filament diameter

3. Gas velocity: The more rapidly a droplet approaches a mesh strand or vane corrugation, the greater its momentum, carrying it in a straighter path. Furthermore, at higher velocities, gas flow streamlines approach the obstacle more closely, resulting in tighter bends. Thus, the capture efficiency of a mist eliminator increases sharply with velocity until an upper limit is reached due to re-entrainment or flooding (discussed later).

4. Liquid density relative to gas density: What causes a droplet to deviate from curving gas streamlines is not its momentum alone, but the difference or ratio between the droplet’s momentum and that of the gas around it. In cases where the gas is nearly as dense as the liquid— for instance, at high pressures—the gas sweeps droplets around the obstacle more strongly, preventing capture.

5. Gas viscosity: The more viscous the gas, the more drag it exerts on suspended droplets as the gas flows around mesh strands and vane corrugations, leading to reduced capture efficiency. The viscosity of a gas generally goes up with higher temperature.

6. Pad density and thickness: Finally, the efficiency of a mesh pad also depends on how closely the strands are packed and on the thickness of the pad. Packing density is increased by knitting with more loops per inch and crimping with narrower ridges. It is measured in terms of pounds per cubic foot of pad. Thickness, in turn, is increased by piling on more layers of crimped mesh sheets. Thicker, denser pads bring trade-offs in terms of higher pressure drop and susceptibility to re-entrainment and flooding. Typical densities for stainless steel mesh are 9 and 12 pounds per cubic feet, and typical thicknesses are 4, 6, and 8 inches.

Interception Capture

There is another capture mechanism, usually called interception, that theoretically applies to both mesh and fiber mist eliminators. (See Figure 14.) Droplets that cannot be captured efficiently by inertial effects due to small size, low density, low velocity, etc., may nevertheless head so close to the centerline of a strand that they brush against the surface and adhere. In practice, however, interception is indistinguishable from inertial impaction and may be ignored in vanes and mesh.

Figure 14: Droplet capture by interception

Brownian Capture

Brownian motion, the main capture mechanism for submicron droplets in fiber mist eliminators, is the frequent random jerks experienced by microscopic particles suspended in a gas or liquid. The cause is momentary inequalities in the number and speed of surrounding molecules hitting the particle from various directions. This tiny motion is enough to throw small droplets out of gas streamlines and against fibers that they would otherwise flow around. (See Figure 15.) Since flow momentum is not involved, capture efficiency is not improved by larger droplets, higher velocity, higher relative liquid density, or lower gas viscosity as for vanes and mesh. Instead, efficiency goes up with higher temperature, longer residence time in the mat (due to greater mat thickness or lower gas velocity), and closer packing of fibers, and down with greater droplet size and pressure.

Because fiber mist eliminators are so different from vane and mesh units in application and specification, further technical information about them is provided in separate Amistco publications.

Figure 15: Droplet capture by Brownian motion in a fiber candle or panel

Capacity Limits

The throughput capacity of a mesh or vane mist eliminator is limited by either of two related phenomena: flooding (choking with liquid) and re-entrainment (dislodging, suspension, and escape of coalesced droplets). In some low-pressure applications, the pressure drop across the device can also be an important consideration.
These limiting factors are illustrated in Figures 16 and 17.

Figure 16 is based on experimental data for a typical horizontal mesh pad (Amistco mesh type TM-1109), using water sprayed at various rates into rising air. It shows how pressure drop varies with velocity and mist load in the vicinity of the typical operating range. The mist droplets are assumed to be within a size range suitable for capture by a pad of this sort—larger than 10 microns.

In Figure 16, notice that the pressure drop would be considered small in most applications—only about 2 or 3 inches of water column even at the most extreme velocity and load combination.

Also notice that pressure drop increases markedly with mist load. At 10 feet per second, the pressure drop for 1 GPM/ft2 is more than three times that for a dry pad.

Figure 16: Pressure drop, flooding and re-entrainment in a typical horizontal mesh pad

Figure 17, in turn, provides a subjective impression of what happens in a typical horizontal mesh pad at three different conditions of flow rate and mist load indicated as Points A, B, and C in Figure 16.

Point A represents a light mist load and a velocity of about 8 feet per second. Nearly all the incoming mist is captured well below the middle of the pad. The rest of the pad remains dry. In the active zone, coalesced droplets slip rapidly down the mesh wire. At the bottom, however, surface tension makes water accumulate on and between wires before falling away as streams and large drops. The result is a thin flooded layer agitated by rising gas, generating a small amount of additional mist that is immediately captured again.

Point B, in turn, lies on a “moderate” load line at the velocity where a few re-entrained droplets begin to blow
upward from the pad—about 11 ft/sec, under these conditions. Re-entrainment is roughly indicated by the darker background at the right side of the plot. (The darker area on the left, in turn, signifies poor capture efficiency.) The higher the liquid load, the lower the velocity at which re-entrainment occurs.

At Point B, velocity is high enough to detach coalesced droplets and lift some of them against the force of gravity. Most re-entrained droplets are relatively large— up to 1,000 microns (1 millimeter). Because of the higher liquid flow rate in the approaching mist and greater upward drag on captured liquid due to higher air velocity, the flooded zone fills an appreciable layer. Incoming mist rises higher in the pad before being captured.

Finally, at Point C, the velocity is high enough not only to lift even the largest re-entrained droplets, but also to retard drainage within the pad virtually to zero. The mesh is entirely choked with agitated liquid, generating mist droplets downstream across a wide range of sizes. Flooding has caused the pressure-drop curve to begin turning up sharply. If flow were increased beyond this point, the line would become almost vertical. For lower liquid loads, flooding occurs at higher velocities.

Similar behavior governs capacity limits also for vane mist eliminators and for horizontal flow through vertical mist eliminators of both types.

As to the influence of operating variables on these phenomena, flooding is promoted by high liquid load (volume percent mist in the incoming mixture), high gas velocity (especially for upward flow as in this example), and high liquid viscosity and surface tension (inhibiting drainage).

At very light liquid loads, re-entrainment can occur without appreciable flooding. However, with or without flooding, re-entrainment is promoted by higher gas velocity, smaller strand diameter or vane corrugation spacing, sharper corrugation angles, greater liquid load, lower liquid density relative to gas, lower liquid surface tension, and lower wettability of the mesh or vane surface.

Figure 17: Envisioning stages in mesh pad performance in preceding figure (vertical cross-sections through pad) (click image to enlarge)

Mesh and Vane Mist Eliminators