new mesh and vane mist eliminator

7/28/2019 New Mesh and Vane Mist Eliminator

1/20

800-231-007714211 Industry Street Houston, TX 77053 TEL: 713-434-0934 FAX: 713-433-6201

Email:[email protected] Visit our website www.amacs.com

The Engineered Mist Eliminator

REDUCECOSTS

INCREASECAPACITY

IMPROVEPERFORMANCE

DEBOTTLENECKEQUIPMENT

SIMPLIFY INSTALLATION

CUSTOMIZEPADS


2/20

The Engineered Mist EliminatorMist elimination, or the removal of entrained liquid droplets from a vaporstream, is one of the most commonly encountered processes regardlessof unit operation. Unfortunately, mist eliminators are often consideredcommodity items and are specified without attention to available tech-nologies and design approaches. The engineered mist eliminator mayreduce liquid carryover by a factor of one hundred or more relative to astandard unit, drop head losses by 50% or more, or increase capacity by

factors of three or four. This manual summarizes input practical approachesto reducing absorbent losses, product contamination and entrainmentcarry over, extending equipment life and maintenance cycles - usingproven and cost effective technologies and techniques.

Droplet Formation and Size DistributionsEntrained liquid does not consist of same-sized droplets, but as a broadrange of droplet sizes that may be characterized with a Normal or BellDistribution centered about some mean or average. The average dropletsize depends very much on the mechanism by which they are generated.Sizing equations are expressed in terms of the probability of removing adroplet of a given diameter, and mist eliminator performance is the

integration or cumulative sum of individual removal efficiencies. It istherefore critical to know the approximate droplet size distribution inorder to properly design a mist elimination system. Figure 1 shows some

typical size distributioncurves from differentsources.

In practice, designersor engineers do notquantify or measuredroplet size distribu-tions, rather they are

assumed based onempirical data or expe-rience. Fortunately, anexperienced engineercan assume anapproximate distribu-tion based on themeans or mechanism

by which the droplets are generated. Typical examples from commonmist sources are given to illustrate these concepts.

Fine droplet distributions, often called fogs (


3/20

equipment corrodes rapidly without the removal of thisliquid. Similar concerns are found in ammonia prilltowers, many chlorine applications, as well asphosphoric and nitric acid plants.

A mist consists of droplets in the range of 3m andgreater, though distributions with average diametersof 20 m and greater are termed Sprays. Mist coming

off the top of packing or trays, or generated by surfaceevaporation, are typically in the broad range of 5-800m.In towers used in glycol dehydration and amine sweet-ening in which mists are a major source of costlysolvent losses, removal of droplets down to 5 m isrecommended.

Hydraulic spray nozzles generate particles with diametersgreater than 50 m and pneumatic nozzles generateparticles with diameters greater than 10 m, with upperlimits reaching 1000 m.

The first step in engineering a mist eliminator is todetermine the mechanism by which the droplets aregenerated and assume an average droplet size.Figure 2 summarizes typical particle size distributionscaused by various mechanisms:

This manual contains basic design concepts used byengineers to remove droplets greater than 3 m in

diameter, so called mists and sprays.

Mechanisms of Droplet RemovalDroplets are removed from a vapor stream through aseries of three stages: collision & adherence to a target,coalescence into larger droplets, and drainage fromthe impingement element. Knowing the size distribu-tions as explained above is important because empir-

ical evidence shows that the target size - important inthe first step of removal - must be in the order ofmagnitude as the particles to be removed. Thesesteps are shown schematically in Figure 3 for mistelimination using a wire mesh mist eliminator.

For fogs in which the bulk of the droplets are charac-terized with submicron diameters, the energy to bringabout the collision with the target is derived fromBrownian Diffusion, the random motion of fine liquidparticles as they arepushed about bymolecular action asshown in Figure 4a.Fog elimination withso-called fiberbedtechnology is beyondthe scope of thismanual.

For particles in themist region between3-20 m, knitted wiremesh is the most com-mon type of misteliminator used andinterception is theprimary mechanism.

THE ENGINEERED MIST ELIMINATOR 2

FIGURE 2

FIGURE 3

DROPLET CAPTURE INAMESHPAD

FIGURE 4

4a

4b

4c

TEL: 800-231-0077 FAX: 713-433-6201 WEB:www.amacs.com EMAIL: [email protected]


4/20

Consider a droplet approaching a mesh filament ofmuch larger diameter as shown in Figure 4b. Themore dense the droplet relative to the gas, the largerthe droplet relative to the filament, and the higher thegas velocity, the more likely it is that the droplet willstrike the filament. If the velocity is too low, or thedroplet too small or too light compared to the gas, thedroplet will simply flow around the filament with the

gas. If the velocity is too high, liquid clinging to the fil-aments will be re-entrained, mostly as larger droplets,and carried away by the gas. Re-entrainment is alsopromoted by low relative liquid density (making it eas-ier for the gas to pick up a droplet) and low liquid sur-face tension (as less energy is required to break up afilm or droplet). The engineered wire mesh mist elimi-nator may remove 99.9% of particles 2 m andgreater diameter. Figure 5 shows a typical removalefficiency vs droplet size distribution for a wire meshmist eliminator.

Droplets ~20 m and greater are primarily collectedby means of Inertial Impaction whereby the target isdirectly in the path of the streamline, as shown inFigure 4c. Figure 6 depicts a profile of the ACSPlate-Pak vane. The entrained droplets, due to their

momentum, tend to movein straight lines. By study-ing this figure, it is easy tounderstand why in thedesign equations to followthe removal efficiency isdirectly proportional to thedifference in densities of

the liquid droplet and carry-ing gas. With each changein direction of the gas,some droplets collide withthe surface and adhere,eventually coalescing intolarger droplets which thendrain by gravity. Properlydesigned vane mist elimi-nators can remove 99% ofparticles as low as 10 min diameter, especially at

lower pressures.

Figure 7 illustrates typicalwire mesh and Plate-Pakvane mist eliminators,and Figure 8 shows some typical performance curvesfor both mesh and vane mist eliminators.


FIGURE 6Dropletcapture in a

Plate-PakTM unit

Stream ofgas curves

back and forthbetween plates

At eachcurve,liquid

dropletsstrikeplates

SEPARATION EFFICIENCY FOR VARIOUSDROPLET SIZES IN A TYPICALWIRE MESH MIST ELIMINATOR

FIGURE 5

FIGURE 7

Typical AMACS

Mist Eliminators

Mesh pad Style 4CA

Plate-Pak unit



5/20

It is worthwhile to discuss Fig. 8 and mist eliminatorperformance. The dotted curves correspond to differentstyles of vanes and the solid to wire mesh styles. Notefirst of all that vanes can be engineered to operate athigher gas velocities and flow rates relative to mesh,but that mesh mist eliminators can approach 100%removal efficiency at smaller droplet sizes. Thisagrees with the discussions above on Interceptionand Inertial Impaction removal mechanisms. Note the

drastic efficiency drop off at low velocities, in whichdroplets drift around the filaments or vane bladeswithout striking them. This phenomenon defines thelower operating range of a mist eliminator. The otherextreme is when the velocity is too high. In this case,the droplets are captured but the velocity of the gasprovides sufficient energy to tear-off and re-entraindroplets. It is in the context of re-entrainment that thedesign equations which follow show that the removalefficiency is directly proportional to the surface tensionof the liquid. As the surface tension increases, so itrequires greater kinetic energy (i.e. gas velocity) to

break the bond between droplet and target, and thedroplets collect and coalesce until drainage by gravi-ty. Re-entrainment defines the upper capacity limit ofa mist eliminator.

Operating range is also affected by the liquid loading(proportion of liquid) of the gas. If too great, the mist

eliminator becomes choked with liquid, a conditioncalled flooding. Flooding is often noticed by highpressure drops or massive carryover of liquids.Typical wire mesh mist eliminators accommodateliquid loads up to about one US GPM per square footand vanes twice as much.

The key operating ranges and suitability of mesh and

vane mist eliminators are summarized in Figure 9. Itemphasizes that vanes are more effective at highervelocities and greater droplet sizes while mesh is moresuitable for removing smaller particles at lower veloci-ties. Gravity settling alone is sufficient for very largeparticles, and co-knit mesh pads, discussed below, forparticles in the range of sizes from 2-8 m. Finally,fiberbed technology is used for submicron fogs.

Types of Mist Eliminator Mesh Styles & MaterialsMost designers believe that all wire mesh mist elimi-nators behave basically the same in terms of capacityand removal efficiency. It is true that for meshes ofsame filament diameter, the denser mesh offers superiorremoval efficiency. For meshes with differing filamentdiameters, a lighter (less dense) mesh may offerconsiderably better removal efficiency. The key is thatthe working part of the mesh is the target density, not


THEORETICAL EFFICIENCY VS. VELOCITY FOR VARIOUSDROPLET SIZES (WATER IN AIR AT AMBIENT CONDITIONS

FOR TYPICAL MESH PADS AND PLATE-PAK UNITSWITH LIGHT LIQUID LOAD)

FIGURE 8

APPROXIMATE OPERATINGRANGES OF MIST ELIMINATORS

FIGURE 9



6/20

the mass density. For example, the most common9-lb density mesh, AMACS style 4CA, exhibits ~85 sq-ft/cu-ft of surface area. Compare this to the co-knit ofa metal with fiberglass (AMACS style 6BE) which alsoexhibits 9-lb mass density but exhibits a specific surfacearea approaching 3,700 sq-ft/cu-ft, some 40X greatertargets per unit volume.

Table 1 shows a few of the more common mesh stylesavailable, together with mesh density and void fraction,and most importantly, the diameter and specificsurface area (i.e. the target density) of filaments used.

It is the amount of targets per unit volume which influ-ences removal efficiency, not the density of mesh(the

greater the number of targets the greater the proba-bility of a successful collision).

In a co-knit such as a metal alloy and fiberglass, thealloy provides a skeleton for structural support andprevents the high specific surface media from collaps-ing on itself.

As far back as the 1950's researchers (C. LeRoyCarpenter et al) determined that specific surface areaand target or filament diameter play a great role inremoval efficiency. Target or filament diameter mustbe on the order of magnitude as the smallest dropletsto be removed. Due to limitations in metal wire ductilityand corrosion considerations, co-knits provide finertargets and hence remove finer droplets. Figures 10

and 11 are enlarged images of crimped wire meshand a co-knit with fiberglass respectively.

In summary, it is important to report mesh styles interms of the specific surface area - a measure of thetarget density, and filament diameter -a measure ofthe smallest droplet size that can be removed withhigh efficiency. The mass density is only relevant insofarthat a metal mesh of density 12-lb exhibits a greaterspecific surface area than one of density 7-lb providedthe wire diameter remains constant.

Selecting the material of mesh style(s) is also important.Corrosion rates as low as 0.005"/year is not serious in

vessel walls but will quickly destroy 0.006" or 0.011"wire mesh. Table 2 gives preliminary guidelines, butAMACS draws wire and knits mesh with any ductile metalfor special applications.

When applying non-metal materials operating temper-ature limits must be considered.


TABLE 1 Wire and Plastic Mesh Styles

CRIMPED WIRE MESH

CO-KNIT MESH WITH FIBERGLASS YARN

FIGURE 10

FIGURE 11



7/20

Design EquationsTo determine mist eliminator cross-sectional area (andhence vessel size) and predict performance in terms ofremoval efficiency, the optimum design gas velocity isdetermined first. The Souders-Brown equation is usedto determine this velocity based on the physical prop-erties of the liquid droplets and carrying vapor:

Vd = k(L-G/G)1/2 (1)

where Vd = design gas velocity (ft/sec)k = Capacity Factor (ft/sec)

L = Liquid Density (lbs/ft3)G = Vapor Density (lbs/ft3)

The capacity factor is determined through experienceand for each application, and is influenced by typeand style of mesh or vane targets used, the geometryof the targets (vertical or horizontal relative to thevapor flow), as well as by properties such as operating

pressure, fluid viscosities, and liquid surface tension.

The design velocity Vd for a given application is thevalue that produces the best performance in terms ofcapturing droplets and avoiding re-entrainment.Referring to Figure 8, this ideal velocity for a givenclass of mist eliminators would be somewhere toward

the upper end of the range: about 10 fps for plain wiremesh pads, about 8.5 fps for co-knits, and 14 fps forPlate-Pak elements. As discussed, effectivenessdrops off at lower velocities as the droplets havesufficiently low momentum to negotiate paths throughthe targets, and at higher velocities because the vaporcarried sufficient kinetic energy to re-entrain droplets.For typical designs, acceptable velocities range

between 25% to 125% of the ideal value.

The Capacity Factor may be thought of as an indicationof ability of a mist eliminator to drain liquids and avoidre-entrainment under various conditions. See Table 3for some typical baseline values.

Note that Souders-Brown equation provides correc-tion for only gas and liquid densities. Should anyconditions exist which affect drainage or re-entrainment,the Capacity Factor must be pro-rated as appropriate.

After selecting the appropriate Capacity Factor andcalculating the ideal vapor velocity, the cross-sectionalarea of mist eliminator is readily determined by dividingthe volumetric flow rate by the velocity.

Having established this design velocity for the appli-cation, you can now predict the efficiency of a mesh

pad for droplets of a particular size. This procedure islaborious and therefore well suited for a computer.

First, calculate the inertial parameter K as follows,using consistent units of measurement:


TABLE 2

Mesh Corrosion & Temp. Considerations

TABLE 3

Standard Souders-Br(k factors) for mesh and Plate-Pak Units



8/20

K = [(L- G)Vd2] / 9D (2)

Where K = dimensionless inertial parameterV = gas velocity in fpsd = Liquid droplet diameter in ft = Gas viscosity in lb/ft secD = Wire or filament diameter in ft

Use this calculated K value with Figure 12 to find thecorresponding value of the impaction efficiency fractionE. From Table 1, find S, the specific surface area forthe mesh style of interest.

Subsequently determine SO of the mist eliminatorperpendicular to vapor flow and with a correction fac-tor of 0.67 to remove that portion of the knitted wirenot perpendicular to the gas flow:

SO = Specific Surface Area x 1/ x Thickness (ft) x 0.67Using these values and T, the thickness of the pad,calculate the capture efficiency:

Efficiency % = 100 - (100/eESO)Where SO = Corrected Pad Specific

Surface Area, ft2/ ft3E = Impaction efficiency fraction

This efficiency is the percent of all incoming dropletsof the given diameter which will be captured ratherthan passing through the mist eliminator. Thepercentage will be higher for larger droplets and lowerfor smaller.

Predicting Pressure DropAlthough the operating pressure differential across aproperly sized mesh pad or vane is never more than afew inches of water, pressure drop is an important

design consideration in certain applications, particu-larly vacuum systems or larger columns requiring themovement of great quantities of gas. It has beenshown that each inch of head loss requires some 0.16hp/scfm. A simple correlation has been developed todescribe the pressure drop through a dry misteliminator (no mist):

Pdry = 0.4VD

2

GST/gc w (3)Where V = Gas Superficial velocity = Ft / Sec

G Gas Density lbs / ft3

S = Specific surface area of mesh ft2/ ft3

T = Mesh Pad Thickness - FtGc = gravitational constant, 32.2 ft / sec2

= Mesh Void Fractionw = Ambient Density of water lbs / ft3

Note: Applicable for wire diameter 0.0045 to 0.015.

The overall pressure drop is the sum of the head lossincurred as the gas travels through the mesh, as well asthat due to the resistance to captured liquids. Liquidaccumulates as a pool in the bottom of the mist eliminator.If the liquid loading and velocity are such that a 2" deeppool accumulates in the bottom of the mesh pad, thisamount must be added to that calculated usingEquation 3. Figure 13 summarizes pressure drop andvelocity test data collected on the AMACSpilot plant for lightand medium liquid loading.

With due consideration given to the mist eliminatoritself, the flow of fluid to and from it requires the sameattention.

Inlet DiffusersAt high flow rates, primary removal of bulk liquidsupstream of the mist eliminator is very important toprevent flooding. This is typically done in a cost effec-tive manner by using a simple inlet diverter as shownin Fig. 15.

With this design, liquids impinge upon the diverters,the flow is forced to flow laterally to allow bulk liquidsto escape by gravity and eliminate the countercurrent

momentum of the gas.The Force of Inertia, expressed as 2, is typicallyused to quantify the flow entering a vessel to determinewhether a simple baffle will suffice. AMACS recommendsinlet diverters to a Force of Inertia up to 2,500 lb/ft s 2.Above this, more sophisticated distributors arerecommended.


DETERMINING IMPACTION EFFICIENCYFRACTION E USING INERTIAL PARAMETER K

FIGURE 12


=


9/20

Decades ago, Dutch Shell Chemical Company intro-duced Schoepentoeter style bladed designs (Fig. 14).

As the fluid flowsaxially towards theshell opposite ofthe inlet nozzle,liquids are cap-tured by speciallyp lac ed blades.

This design issuperior because itallows the escapeof liquids over amuch g rea te r

region of the vessel. A simple inlet diverter ( Fig. 15)would simply shear bulk liquids into smaller dropletsat great flow rates:

AMACS AccuFlow

Inlet Diffuser (Fig.16) is a similiar

style of the bladeddesign in whichthe body of the dif-fuser maintains itsshape, the restric-tion of flow whichallows the escape

of liquids over thediameter of thevessel is accom-plished using inter-nal blades of con-centric and decreas-ing cross-sectionalareas.

Vessel ConfigurationSeveral factors must be considered when deciding onthe configuration of vessel internals. The first step isto determine the cross-sectional area needed. Then atentative geometry and shape appropriate for both thevessel and plant location is selected. Figure 17 showsthe most typical, but by no means complete, configu-rations. Mist eliminators can be of virtually any size orshape to accommodate all factors.

The performance of the mist eliminator dependsstrongly on an even velocity distribution over thecross-sectional area. As a general rule, a distance ofeither half the vessel diameter or 72", which ever issmaller, is sufficient spacing both upstream and down-stream of the element. Representations for specificcases are illustrated in Figure 18.


ACTUAL PRESSURE DROP VERSUS VELOCITY FOR TYPICAL AMACS MESH PADS AT LIGHT AND MEDIUM LOADS

FIGURE 13

FIGURE 14

FIGURE 15

FIGURE 16



10/20

Small velocity differences across the surface areacceptable, but should be minimized at the designstage. Otherwise, some regions of the mist eliminatormay be subjected to heavy loading leading to re-entrainment while other regions are unused.

Most often, the mist eliminator is located justupstream of the outlet nozzle with insufficient disen-gagement space. Vapor tends to channel through the

pad in the region closest to the outlet nozzle andperipheral regions of the pad remain unused. To rectifythis, AMACS engineers apply an Integral Flow Distributorwhich is welded to region(s) of the downstream faceof the pad. This technique allows the engineer toselectively increase the pressure drop throughregions of the pad likely to suffer from channeling, andis cost effective.


SIMPLIFIED VIEWS OF TYPICAL MIST ELIMINATOR CONFIGURATIONS IN SEPARATOR VESSELS

FIGURE 17



11/20

Advanced Mist Eliminator DesignsThere are several modifications to mesh pads andvanes to dramatically enhance performance.

Drainage & Collection LayeringRecall the discussion on pressure drop through a misteliminator in which liquid tends to pool in the lower layersof mesh. The simplest technique to promote drainageis to use a few inches of open, porous mesh such asAMACS style 7CA (5-lb density with specific surface areaas low as 45 sq-ft/cu-ft) in the upstream position. Asdrainage occurs through the interstitial regions of the

mesh, opening theknit enhances liquiddrainage.

An extension of thisapproach is to usehigher specific surfacearea mesh in down-stream positions toenhance separationefficiency, withintermediate mesh

between the collection and drainage zones. Figure 19illustrates a multilayer mist eliminator.

MisterMesh Drainage CoilsA second technique used by AMACS to enhance liquiddrainage, and often in conjunction with multi-layering,is to append drainage coils to the upstream face of ahorizontal mist eliminator as shown in Figure 20.The coils are also made of mesh and "fill" with liquid.

Once filled, liquid from the pad above is drawn bygravity and The Coanda Effect to the coils, therebyestablishing distinct regions for liquid drainage and

liquid collection in the upstream layers. Figure 21compares the pressure drop and flooding point ofboth conventional and MisterMesh Mist Eliminators.


Guidelines for maintaining even flow distribution acrossmesh pads or vane units with axial flow in cylindrical ves-sels. Height of vessel head is assumed to be 1/4 of vessel

diameter. Flow distribution devices can minimize requireddisengagement space above mesh pads.Contact AMACS for assistance.

FIGURE 18

FIGURE 19

MISTERMESH PAD WITH DRAINAGE ROLLS

FIGURE 20



12/20

Mesh-Vane AssembliesIn grass root design of larger vessels and retrofit ofexisting ones to accommodate greater flow rates, mesh-vane assemblies are often used. In an assembly, meshis placed upstream of the vane and acts as a floodedagglomerator. The capacity factor used corresponds tothe downstream vane element. This approach com-bines the efficiency of mesh with the capacity of vanes

and has been used by AMACS engineers with tremendoussuccess over the past two decades.

Throughout the industry there is ongoing debate as towhether the mesh should be positioned up- or down-stream of the vane element. Engineers at AMACS haveperformed exhaustive comparative testing on pilotplants and have much field data proving that the meshis indeed affective upstream of the vane, unless the

vane element is used as a pre-filter to protect a down-stream mesh pad.

Use of GeometryAnother approach used in the industry when the sizeof the vessel is limited is to arrange the mist eliminator atan angle. The capacity increase is equal to the sine ofthe angle though it should not exceed 45. This is

shown in Figure 22 for smaller and larger diameters.An AMACS engineer should be consulted for such designs.

MultiPocket VanesThe capacity of vertical vanes (with horizontal vapor

flow) can also be increased by enhancing liquiddrainage. As discussed, captured liquids are re-entrainedwhen the velocity of vapor exceeds the ideal. Toprevent liquid re-entrainment, the serpentine pathoffered by the vane is augmented with obstructions toallow for the pooling of liquid with protection from thepassing vapor stream. This design increases thecapacity of the vane by as much as 25%. In verticalgas compressor knock-out drums, in which the vesselsize is dictated by the capacity of the mist elimina-tor, MultiPocket

Vanes considerablyreduce the Foot-printand cost of skids.

Figure 23 summarizesthe approaches usedby AMACS and thereduction in vesseldimensions possibleusing these advanceddesigns.

The MultiPocket

Vanehas been patented byAMACS.


ACTUAL PRESSURE-DROP PERFORMANCEOF MESH PADS VERSUS VELOCITY.NOTE RAPID INCREASE AS FLOODED

CONDITION IS APPROACHED

FIGURE 21

FIGURE 22

FIGURE 23



13/20


(Now AMACS Process Tower Internals)



14/20



Amistco Separation Products Inc, and ACS Separations and Mass Transfer Products, are Registered Trademarks of AMACS

MistFix Insertion Mist Eliminators Mist Eliminators

Advantages:

No Cutting of existing vessel

No Welding

No Hazardous Entry

No ASME re-certification

No Scaffolding

Minimal Downtime

MistFix U.S. Patent #5985004

For more information please call:

1-800-231-0077www.amacs.com

.

The patented AMACS MistFix can solve carryover problems in

vessels without a mist eliminator, as well as in vessels with a less

efficient or damaged mist eliminator.

In existing vessels that do not have a manway, the MistFix Insertion

Mist Eliminator is an ideal choice. It is suitable for any vessel having

an 8 or larger gas outlet nozzle at the top. It also eliminates the need

for hazardous entry permits. Since there is no need to enter the

vessel, this drastically reduces downtime, resulting in quicker

turnarounds, reduced maintenance cost and production gains.

MistFix also eliminates the need for modifications to vessels. Fornew vessels MistFix may eliminate the need for a manway and

reduce vessel cost. It also makes future maintenance easier and

simpler.

AMACS MistFix can easily be installed and replaced from the

outside. Existing vessels require no modifications to accommodate the

MistFix.

Figure 3. MistFix insertion mist eliminator


15/20

Our

are registered trademarks

of AMACS Process Tower Internals.

14

Try AMACS Plate Pak vane

our

we


16/20

CASE STUDIES & EXAMPLESCase Study Number 1Problem: In an HCl scrubber, an air stream of 60 acfsis coming off a bed of random packing and containsdroplets of a weak acid. The unit operates at 122 psiaat 82F. Determine the size of mist eliminator requiredto remove this mist and the removal efficiency possible.

Solution: Since the acid is dilute we assume the densityand viscosity of water at the operating pressure andtemperature:

L = 62.4 lb/ ft3

G = 0.60 lb/ ft3

P = 122 lb/ft2

T = 82FF = 60 ft3/sec

The first step is to select the mist eliminator type andmesh style. As shown in Figure 24, mist coming to themesh pad is typically comprised of droplets ranging insize from as small as 5 m, so we select a mesh stylemist eliminator to achieve this level of performance. Fromexperience, the capacity factor for poly mesh at mod-erate liquid loading and lower pressures is ~.27fps. Using the Souders-Brown equation the idealvelocity is calculated:

Videal = k [ (L- G) /G]1/2

Videal = 0.27[(62.4-0.60)/0.60]1/2

Videal = 2.74 fpsThe cross-sectional area of mist eliminator is deter-mined by dividing the volumetric flow rate by the idealvelocity:

Area Mist Eliminator = Volumetric Flow Rate/Superficial Vapor Velocity

Area Mist Eliminator = [60 ft3/sec]/2.74 fps

Area Mist Eliminator = 21.9 ft2

The corresponding diameter is 63.4", rounded up to a

standard 66" scrubber vessel. Note that performingthe same calculations using a vane (and a capacityfactor of 0.50) yields an ideal vessel diameter of 46.7",rounded up to a standard 48" ID vessel. To calculatethe removal efficiency at 5 m, several parametersmust be identified to use equation 2 to determine theinertial parameter K:


APPLYING COMBINATIONS OF AMACS MESH PADSAND PLATE-PAKTMVANE UNITS TO MINIMIZE VESSEL

SIZE

FIGURE 24

TEL: 800-231-0077 FAX:713-433-6201 WEB:www.amacs.com EMAIL: [email protected]


17/20

K = [(L- G)Vd2]/9D

K = 0.32 fps

From Figure 12, the corresponding ImpactionEfficiency Fraction E is ~0.08. In the RemovalEfficiency Equation there is a term for the correctedspecific surface area SO:

SO= Specific Surface Area x 1/ xThickness (ft) x 0.67

For ACS style 8P, the specific surface area is(185 + 36) = 221 ft2/ft3, we will try both 4" and 6" thickmist eliminator thicknesses (1/3 and 1/2ft):

SO = 221 x 1/3.14 x 1/3 x 0.67SO 4"thick = 15.7 and SO 6"thick = 23.6

And Removal Efficiency E at 5 m is:

Efficiency = 100 100/eESO

Efficiency = 100 100/e(0.08)(15.7)Efficiency = 71.5%

For the 6" thick element, the removal efficiency is84.8%. By using a composite pad containing a 2"layer of regular monofilament polypropylene, style 8P,upstream of a 2" thick layer of 8PP, mono- andmulti-filament co-knit, the removal efficiency is 99.9% .

CASE STUDY #2Traditionally, trays are used to bring about contact

between glycol and natural gas in dehydration con-tactors. In recent years, the industry moved towardssmaller diameter columns by exploiting the highercapacities achieved with structured packing.However, the lower capital investment associated witha smaller diameter packed tower is often offset bydramatically increased glycol losses.

Consider a mid-western sour gas plant operating a96" glycol contactor and processing 1,310,000 lb/hr ofgas at 116F and 1214 psia. The gas and liquid specificdensities were 4.4 and 68 lb/cu-ft respectively. The

plant was experiencing 0.13 US gal of carryover permmscf, amounting to some 65 gal/day of lost triethyleneglycol, several hundred dollars worth per day. A 10"thick wire mesh mist eliminator of 12-lb mass densitywas installed above the packing.

From experience, AMACS engineers knew that thedroplet size distribution for glycol coming off the top ofa packed dehydrator extends down to diametersof 5 m and greater. Also, if the diameter of thepacked column was sized in accordance with thehydraulic requirements of the packing, the wire meshmist eliminator would be undersized.

The capacity factor for 12-lb density mesh in this serv-ice is ~0.23 0.27, having been de-rated for the highliquid viscosity of 18 cP (which retards liquid drainage)and relatively high operating pressure. Using the gasdensity, volumetric flow rate and cross-sectional areaof the mist eliminator, the actual superficial velocity isreadily calculated. Next, using known densities of thegas and glycol, the actual or operating CapacityFactor k is determined:

Vactual = kactual [(L- G) /G] 1/2

Re-arranging forkactual = Vactual/ [(L- G) /G] 1/2

= 0.44 fps

A Capacity Factor of 0.44 fps is almost twice as highas the optimum, and is in the range of that of an AMACSPlate-Pak Vane mist eliminator. However, the vanewill not remove particles down to 5 m, so a mesh-vane assembly was proposed. The assembly has amultiple layers of mesh. The first layer is composed ofhighly porous mesh (AMACS style 7CA), followed by alayer of the high specific surface area (AMACS style 8DT)co-knit mesh of stainless and Dacron Fibers.MisterMesh drainage coils were appended to thebottom face of the mist eliminator. Downstream of themesh was placed a Plate-Pak vane. The totalthickness was 12" and was accommodated using thesame supports as the mist eliminator it replaced.

Carryover from a glycol contactor occurs through twomechanisms, evaporative losses and mechanical

(carryover losses). In this example, simulationsshowed evaporative glycol losses of 0.0054gal/mmscfd. The total losses after the revamp wereless than 0.008 gal/mmscfd, and carryover losses hadbeen reduced from 0.13 gal/mmscfd, a 94% reduction!


TEL: 800-231-0077 FAX:713-433-6201 WEB:www.amacs.com EMAIL: [email protected]


18/20



19/20



20/20

24-hour emergency service Free technical support 50 years experien

Our MisterMesh Mist Eliminator out per-forms conventional pads. The drainage roaccelerate liquid removal thus increasingcapacity and reducing pressure drop. Uin conjunction with our Plate-Pak van

the MisterMesh drainrolls can increase capacby over 200% while seprating droplets down t3 microns.

PLATE-PAKM IST ELIMINATOR

M ISTER MESH

M IST ELIMINATORWITH DRAINAGE ROLLS

ADVANCED TECHNOLOGYFOR DEBOTTLENECKING !

OUT PERFORM CONVENTIONALTECHNOLOGY!HIGH CAPACITY MIST ELIMINATORS

800-231-007714211 Industry Street Houston, TX 77053 TEL: 713-434-0934 FAX: 713-433-6201

Email: amacs@amacs com Visit our web site www amacs com

new mesh and vane mist eliminator

Documents