DP03B

ExxonMobil ProprietaryFRACTIONATING TOWERS Section Page

SIEVE TRAYS III-B 1 of 83

DESIGN PRACTICES December, 2003

ExxonMobil Research and Engineering Company – Fairfax, VA

CONTENTS

Section Page

SCOPE............................................................................................................................................................ 4

REFERENCES................................................................................................................................................ 4DESIGN PRACTICES............................................................................................................................. 4GLOBAL PRACTICES............................................................................................................................ 4OTHER REFERENCES.......................................................................................................................... 4

BACKGROUND .............................................................................................................................................. 5

DEFINITIONS / EQUATIONS ......................................................................................................................... 5

APPLICATION ................................................................................................................................................ 5

BASIC DESIGN CONSIDERATIONS ............................................................................................................. 6TOWER DIAMETER............................................................................................................................... 6TRAY SPACING ..................................................................................................................................... 6NUMBER OF LIQUID PASSES .............................................................................................................. 7

Transitions ........................................................................................................................................... 7TRAY AND DOWNCOMER LAYOUT..................................................................................................... 7

Hole Area............................................................................................................................................. 7Hole Size ............................................................................................................................................. 8Blanking ............................................................................................................................................... 8Downcomer Width And Area................................................................................................................ 8Outlet Weirs And Downcomer Clearances........................................................................................... 8Tray Balancing..................................................................................................................................... 8Multipass Trays.................................................................................................................................... 9Multipass Tray Balancing..................................................................................................................... 9Column Access.................................................................................................................................... 9Startup Considerations ...................................................................................................................... 10

PROCESS CONSIDERATIONS ........................................................................................................... 10Tray Turndown And Weeping ............................................................................................................ 10Tray Efficiency And Heat Transfer ..................................................................................................... 11Low Liquid Rate Tray Design............................................................................................................. 11High Liquid Rate Tray Design ............................................................................................................ 11Foaming............................................................................................................................................. 11Vapor Recycling................................................................................................................................. 12Fouling ............................................................................................................................................... 12Corrosion ........................................................................................................................................... 13

TOWER CHECKLIST ........................................................................................................................... 13

CAPACITY/ PERFORMANCE RESTRICTION MECHANISMS ................................................................... 13OVERALL CAPACITY .......................................................................................................................... 13

Overall Flood ..................................................................................................................................... 13Probability Of Non-Flooding Operation .............................................................................................. 13

VAPOR HANDLING LIMITATIONS ...................................................................................................... 14Jet Flooding ....................................................................................................................................... 14

ExxonMobil ProprietarySection Page FRACTIONATING TOWERS

III-B 2 of 83 SIEVE TRAYSDecember, 2003

Ultimate Capacity ...............................................................................................................................15Spray Regime And Entrainment .........................................................................................................17

LIQUID HANDLING LIMITATIONS........................................................................................................20Downcomer Flood ..............................................................................................................................20

SECONDARY DESIGN PARAMETERS ...............................................................................................23

SIEVE TRAY DESIGN PROCEDURE ...........................................................................................................26EMOTIP DESIGN ALGORITHM............................................................................................................26AVAILABLE PROGRAMS .....................................................................................................................32

TABLES

Table 1 Sieve Tray Design Principles ...........................................................................................................33

Table 2 System Factors ................................................................................................................................35

Table 3 Equations For Determining Liquid And Vapor Splits ........................................................................38

Table 4 Default Design Algorithm Values......................................................................................................40

FIGURES

Figure 1 Weeping And Dumping Regions .....................................................................................................44

Figure 2 EMoTip Tray Performance Diagrams.............................................................................................45

Figure 3 E-Method Entrainment Kφ Factor ....................................................................................................48

Figure 4 E-Method Entrainment Kl Factor.....................................................................................................49

Figure 5 E-Method Entrainment Kσ Factor....................................................................................................51

Figure 6 E-Method Entrainment Kε Factor ....................................................................................................52

Figure 7 Kσµ Factor For E-Method Entrainment Correlation .........................................................................54

Figure 8 Three-Pass Tray Geometry.............................................................................................................55

Figure 9 Four-Pass Tray Geometry...............................................................................................................56

Figure 10 EMoTip Sieve And Valve Tray Design Algorithm..........................................................................57

Figure 11 Dry Tray Pressure Drop Design Consideration Function ..............................................................58

Figure 12 Liquid Load Design Consideration Function .................................................................................59

Figure 13 Froth/Spray Transition Design Consideration Function.................................................................60

Figure 14 Downcomer Choke Design Consideration Function......................................................................61

Figure 15 Entrainment Design Consideration Function.................................................................................61

Figure 16 Flow Path Length Design Consideration Function ........................................................................62

Figure 17 Weeping Less Than 20% @ Turndown ........................................................................................63

Figure 18 Weeping Rate Design Consideration Function .............................................................................64

Figure 19 Sealing Factor @ Design Rates Design Consideration Function..................................................64

Figure 20 Sealing Factor @ Turndown Rates Design Consideration Function ..............................................65

Figure 21 Vapor Carryunder Design Consideration Function........................................................................65

Revision Memo

Revision marks are not included in this revision because this is essentially a complete rewrite.

12/03 The highlights of this revision are:

DESIGN PRACTICES

1. The Jet Flood, Ultimate Capacity, Probability of non-flood, Weeping, and TABLE 2 "Design Criteria"have been replaced with EMoTIP correlations and/or values.

2. Overall Flood has been added.3. Downcomer Flood has been added.4. Foaming Factor and Fouling Factor have been added.5. Froth to Spray Transition correlation design limits have been modified.6. Downcomer Seal correlations have been modified.7. M-Method Entrainment correlation has been added.8. Universal Ultimate Capacity correlation has been added.9. Tray balancing explanation has been modified.10. New design limits have been added.11. Deleted all Figures in the Appendix that represented old correlations.12. Deleted SIEVE TRAY CALCULATION FORMS at the end of the Section.13. Added clear liquid height term used in deck frothing, as well as clear liquid height term used in the total tray

pressure drop and downcomer backup calculations.14. Added a discussion of the EMoTIP design algorithm.15. Updated discussion of general design considerations.16. Added figure showing tray performance diagram calculated with EMoTIP for three pressure levels.17. Added figures showing the design consideration weighting function for the various design variables.18. Added the EMoTIP weep point and weep rate correlation.19. Changed the "Detailed Design Procedure" section to a "Capacity/Performance Restriction Mechanism"

section to reflect the EMoTIP approach.20. Restructured the "Basic Design Considerations" section to better reflect the EMoTIP approach.21. Mention of old programs have been replaced with EMoTIP.22. Updated TABLE 3 EQUATIONS FOR DETRMINING LIQUID AND VAPOR SPLITS

SCOPEThis section covers the techniques for specifying the process design features of sieve trays for new designs or revamps. It isassumed that the designer has already read Section III-A, Device Selection and Basic Concepts, and determined that sieve traysare the best choice for the design. The ExxonMobil Tower Internals Program (EMoTIP) utilizes the equations and criteriapresented in this section for new tray designs and for rating existing trays. A discussion of the ExxonMobil Tower InternalsProgram (EMoTIP) design algorithm is included in this section. Detailed mechanical design as well as beam and hole layout arenormally handled by the tray fabricator and therefore are not discussed in this section. A list of FRACTIONATION SPECIALISTSto contact for help is provided at the beginning of Section lII.For the design of tray-related tower internals, such as nozzles, drawoff boxes and reboiler connections, refer to Section lII-H,Tower Internals. For the design of heat transfer trays, see Section III-F. To calculate tray efficiency, see Section lIl-l. Areas andlengths of chords are given in Section III-K.

REFERENCES

DESIGN PRACTICESSection III, Fractionating Towers

GLOBAL PRACTICESGP 05-02-01, Internals for Towers, Drums and Fixed Bed Reactors

OTHER REFERENCES1. Becker, P. W. and Peruyero, J. M. A., Minimizing Entrainment in Sieve Tray Towers, ER&E Report No. EE.64E.77, June,

1977.2. Colwell, C. J., Low Liquid Rate Entrainment on Sieve Trays, ER&E Memorandum No. 83CET 45, January 11, 1983.3. Kaplan, R. H., New Correlation Predicts Froth to Spray Transition on Sieve Trays, ER&E Report No. EE.128E.82,

December, 1982.7. Stober, B. K., NDG Extractive Distillation Tower (T-1420) Performance Tests and Tower Internal Revamp, ER&E

Memorandum No. 88 CET 123, April 4, 1988.8. Wood, S. M. and Stober, B.K., Evaluation of Sieve Tray Capacity Correlations, EMRE Report No. EE.76E.2003, April, 2003.9. Stober, B. K., Tower Internals Design Memorandum No. 1: Recommended Ultimate Capacity Correlation for Use with

Packing or Trays, EDSFile: T-FRA-PACK/TRAY, January 26, 1990.10. Stober, B. K., Tower Internals Design Memorandum No. 2: Use of New Flooding Correlations for All Tray Designs, EDS File:

T-TWINT-CAP, * January 23, 1991.11. Chern, J. E. and Stober, B. K., Tower Internals Design Memorandum No. 3: Development of Mobil Overall Flood, EDS File:

T-TWINT-CAP,* September 21, 1992.12. Buchanan, J. S., Tower Internals Design Memorandum No. 5: Improved Correlation for Sieve Tray Turndown, EDS File: T-

TWINT-FLUID FLOW,* December 5, 1995.13. Buchanan, J. S. and Grave, E. J., Tower Internals Design Memorandum No. 11: Effects of High Liquid Viscosities on

Packing and Tray Capacities, DAN: 98M-0623,* June 23, 1998.14. Buchanan, J. S. and Nguyen, H-T. D., Tower Internals Design Memorandum No. 12: Revised MoTIP Jet Flood Correlation,

DAN: 98M-0650, * July 1, 1998.15. FRI Topical Reports: 88 Pressure Drop of Sieve Trays, December 1982; 101 Model for Downcomer Flooding of Sieve Trays,

September 1986; 119 Models for Liquid Head, pressure Drop and Weeping of Sieve Trays, October 1995.16. Stober, B. K., EMoTIP Sieve Tray Hydraulics Equations, EMR&E Memorandum No. 2003 APTD 7, March 17, 2003.17. Guarda, C. F., Design Practices Section III-B Sieve Trays 1999, EMR&E Memorandum No. 2003 APTD 14, March 14, 2003.

* Tower Internal Design Memorandum No. 1-15 have been archived in electronic form under 2003 APTD 121.

DESIGN PRACTICES

BACKGROUNDThe equations presented in this section for calculating sieve tray capacity and hydraulics are either ExxonMobil developed models orFractionation Research, Inc. developed models that have been modified by ExxonMobil to improve the fit to the availableFractionation Research, Inc., (FRI) data and data from simulator and commercial tests. These equations represent the overall datamore accurately than the correlations prepared by FRI, various vendors, or those available in the literature. These equationssupercede those used in the Sieve Tray Design Program 1133 and the Multipass Sieve Tray Design Program 1143.

DEFINITIONS / EQUATIONSFor a discussion of such concepts as weeping, dumping, spray regime transition, jet flooding, downcomer flooding, overall flood,choking, efficiency, entrainment, flexibility, etc., see Section III-A, Device Selection and Basic Concepts. See NOMENCLATURE atthe end of this section for symbol definitions.Because of the complexity of the new models described in this section, they are no longer appropriate for hand calculation.Therefore, the SIEVE TRAY CALCULATION FORMS have been deleted from this revision. For those engineers who need to referto the previous version of this Design Practice Section to access the SIEVE TRAY CALCULATION FORMS or to check the 1133 or1143 program results, it has been archived as Reference 17. It can be accessed through the ExxonMobil eMemory application. Also,in some cases full details of the hydraulic models are not presented in this section. Those engineers who would like full details of thehydraulic models should retrieve Reference 16 which contains all the equations necessary to hydraulically rate a standard single passsieve tray. The ExxonMobil Tower Internals Program (EMoTIP) incorporates all the calculations discussed in this section and is therecommended ExxonMobil tool for designing and rating sieve trays.Equations that do appear in this version have been renumbered. They are presented in a form to calculate one pass of a tray only.By and large, only the customary unit versions of the equations are presented in the text. The main purpose for presenting anyequations in this Design Practice is to give the design engineer an understanding of the functional forms and, where possible, theeffects of the various design parameters on the hydraulics of a sieve tray. The equations presented here have been validated incustomary units only and are not recommended for hand calculations. EMoTIP should be used for all calculations and fordesign purposes.

APPLICATIONSieve trays can be used in almost all services. Their capacity and efficiency are at least as high as that of other standard trays usedcommercially. Flexibility is generally around 2/1, but ranges up to a maximum of about 3/1. For greater than 3/1 flexibility, valve traysare a better choice.

Sieve trays may be used in moderately fouling services, provided that large holes (3/4 to 1 in. [19 to 25 mm]) are used.

The following table lists the lower and upper operating limits based on the database used to develop the correlations and operatingexperience. This table contains the current limits over which the correlations contained in this design practice are considered to beaccurate. If your case does not fall within these limits, contact your FRACTIONATION SPECIALIST to see what, if any, problemsmay exist. These are not recommended design values, for that see Table 1 and the other information contained in this section.

VARIABLE LOWER LIMIT UPPER LIMIT

Pressure, psia (kPa) 3 (21) 450 (3100) distillation900 (6200) absorption

Temperature, °F (°C) 130 ( 90) 800 (430)Diameter, ft (mm) 1.0 (300) 50 (15,240)Physical properties

surface tension, dyne/cm (mN/m)liquid viscosity, cP (mPa•s)vapor density, lb/ft3 (kg/m3)liquid density, lb/ft3 (kg/m3)

1 (1)0.05 (.05)

0.005 (.08)20 (320)

75 (75)20 (20)5 (80)

80 (1300)Tray spacing, in. (mm) 12 (300) 36 (910)Open Area as % of Ab 3.5% 15%Downcomer clearance, in. (mm) 1 (25) 3.5 (90)Downcomer inlet area as % of As 6% 40% sloped; 25% straightNumber of passes 1 4Outlet weir height, in. (mm) 0 (0) 4 (100)Hole diameter, in. (mm) 1/8 (3) 1 (25)Flow path length, in. (mm) 16 (410) for access 180 (4600 mm)

BASIC DESIGN CONSIDERATIONS

The ExxonMobil Tower Internals Program (EMoTIP) is available for designing and rating trays for fractionation columns.However, before using EMoTIP, it is essential that the designer have a basic understanding of the key parameters that influencetray design. This section provides a discussion of these key parameters and presents most of the equations used by the EMoTIPto calculate them. This section also includes certain "rules of thumb" that can aid the designer in achieving an optimum traydesign.

The optimum combination of tower diameter, tray spacing, and number of liquid passes is the most significant consideration innew designs affecting tower cost and maintenance.

TOWER DIAMETERSee the above table for the limits on the minimum and maximum tower diameter when using this Design Practice. AFRACTIONATION SPECIALIST should be consulted on tower designs outside these limits. The tower diameter must provideenough cross-sectional area to avoid downcomer flood, jet flood, and ultimate capacity limitations. Large towers are sometimesdesigned in sections, with each section having a different diameter. This practice is not suggested for small towers.

TRAY SPACINGTray spacing is normally set to allow easy access for maintenance. A tray spacing of 24 in. (610 mm) is the most common forcolumns 4 ft (1219 mm) and larger in diameter. This spacing is large enough to allow a worker to freely crawl between trays. Forcolumns where frequent maintenance is expected, such as fouling and corrosive services, a tray spacing of at least 24 in. (610mm) is recommended. A tray spacing of at least 24 in (610 mm) is also preferred for systems with a high foaming tendency. Forcolumns smaller than 4 ft (1219 mm), a tray spacing of 18 in. (457 mm) is adequate for maintenance. Here, crawling betweentrays is uncommon because a worker can reach the column wall from the manway. A tray spacing smaller than 18 in. (457 mm)should be avoided because it makes access for maintenance difficult. However, in columns containing 100 - 200 trays, such asC2/C3 splitters, tray spacing can be as low as 12 to 18 in. (305 to 457 mm) to prevent excessive column height. Downcomer floodand jet flood requirements may require the use of tray spacings larger than the minimum. Spacings up to 36 in. (900 mm) may beused to permit a higher superficial vapor velocity or downcomer flood. While the ExxonMobil Tower Internals Program (EMoTIP)design algorithm selects tray spacings at 3 in. (75 mm) intervals for convenience, the designer is free to use any tray spacingdesired as long as it is within the acceptable range of 18 to 36 in. (457 to 900 mm). The following table gives the minimumrecommended tray spacing values determined by maintenance considerations and support beam depth.

DESIGN PRACTICES

MINIMUM RECOMMENDED TRAY SPACING

FOULING SERVICE (FOULING FACTOR > 1)CLEAN SERVICE

(FOULING FACTOR = 0 OR 1) 1-Pass 2 or More Passes

TOWER DIAMETER, ft(mm)

in. Mm in. mm in. mm

5 or less (≤ 1500) 18* 457 18* 457

5 to 8 (1500 to 2400) 18* 457 24 610 18* 457

8 to 10 (2400 to 3000) 18* 457 24 610 24 610

10 to 20 (3000 to 6000) 18* 457 24 610 24 610

Greater than 20 (> 6000)** 24 610 27 686 27 686

Notes:

* If there is no manhole between trays. Minimum tray spacing with a manhole is 24 in.

** For towers larger than 20 ft (> 6000 mm) in diameter, "lattice" type trusses must be used to facilitate maintenance and permit good vapordistribution. (See Section III-H for a picture of a lattice truss.)

NUMBER OF LIQUID PASSESThe capacity of towers with high liquid rates can usually be increased by the use of multipass trays. Since multipass traysincrease the sensitivity to maldistribution, which may result in decreased efficiency, and are more expensive than single passtrays, they can be justified only if their use reduces the overall tower cost. Generally, this means that a capacity advantage of atleast 5 to 10% for multipass trays is required. However, each case must be studied on its own merits, since overall tower costdepends on many factors, including height, diameter, operating pressure and materials of construction. If the liquid rate is greaterthan 17.5 gpm/in. of outboard weir/pass (43.5 dm3/s/m of outboard weir/pass), a FRACTIONATION SPECIALIST should beconsulted because of the lack of reliable design data above this rate. More detailed selection criteria are given in Table 1.If an existing tower is limited by downcomer flooding, which cannot be reduced by other hardware changes, the use of multipasstrays should also be considered.If a two pass design can be found, it will generally be preferred over a four pass design, due to increased cost for four passdesigns and increased risk due to tray balancing and installation tolerances that are critical for four pass designs.

TransitionsChangeover from one number of liquid passes to another is frequently required where a feed stream or a circulating reflux streamis introduced. It is important to verify that such transitions do not restrict flow, cause maldistribution, or result in downcomersealing problems. One to two pass transitions and two to four pass transitions are the most common transitions. This is becauserectifying sections tend to have lower liquid rates than stripping sections and therefore require fewer passes. Refer to Section III-H for methods for achieving successful transitions from one number of liquid passes to another.

TRAY AND DOWNCOMER LAYOUTTwo important features of the tray layout are the bubble area Ab and the free area Af (see Figures 12 and 13 in Section III-A).These in turn, depend on the liquid handling areas (downcomers) and waste area Aw, defined as any unperforated area fartherthan 3 in. (75 mm) from the edge of the nearest perforation. Normally, there is no waste area on a sieve tray unless a very lowhole area is required (part of the tray is left unperforated) or if a shaped downcomer lip, recessed inlet box, or inlet weir is present.

Hole Area

The hole area on the tray should be large enough to avoid operating in the spray regime and small enough to ensure thatexcessive weeping is avoided. Hole area has a direct effect on dry tray pressure drop. The only way to obtain the desired drytray pressure drop is to adjust the hole area. Increased hole area also helps reduce downcomer flood by reducing downcomerbackup.

Hole Size

Hole diameters on sieve trays usually range from 1/8 in (3 mm) to 1 in (25 mm). For most cases, a hole size of 1/2 in. (13 mm)should be used. The allowable range of hole sizes is outlined in Table 1. Determining the appropriate hole size for sieve traysdepends on various factors such as: the nature of the service, tray hydraulics, and turndown. Small holes are not recommendedfor fouling or corrosive services because they may plug or partially plug the orifices on the tray, resulting in excessive pressuredrops, decreased capacity, and lower efficiency. However, smaller holes increase jet flood capacity, particularly at low liquidloadings when operating in the spray regime. They will reduce the entrainment rate and reduce slightly the dry tray pressure drop.Small holes have a higher weeping tendency. While R&D studies have indicated that smaller hole sizes (0.125 in. [3 mm]) dohave better entrainment characteristics for some systems, small holes on carbon steel trays have a tendency to "rust over" duringtower hydrotesting or storage. 410 SS trays should always be compared with carbon steel, because their favorable corrosioncharacteristics for most refinery applications means that the thinner 410 SS decks are cost competitive with thicker carbon steeldecks that include a corrosion allowance. There is a small but distinct process performance advantage for thinner trays. Formoderately fouling services, hole sizes of 3/4 to 1 in. (19 to 25 mm) are recommended. Sieve trays are not recommended forhighly fouling services. Do not mix panels with different hole sizes on the same tray.

Blanking

For revamps, hole area may be reduced by either using blanking strips or replacing the panels with ones having a smaller holearea. If blanking 50% or less of the hole area, blank single rows or pairs of rows of holes. If pairs of rows are blanked, leavesingle rows open; conversely, if single rows are blanked, leave pairs of rows unblanked. This minimizes the channeling of frothover the blanked tray. Blanking patterns should begin with either the second or third row adjacent to the outlet weir (depending onwhich blanking pattern is chosen), and shall proceed towards the inlet side of the tray. Blanking strips must always beperpendicular to the froth flow on the tray.For large amounts of blanking (50% of the hole area or more):• Use a combination of items mentioned above, but make sure that all panels have the same effective hole to bubble area ratio.

Otherwise, channeling may result.• Consider adding vertical baffles to restrict flow path width and create what is commonly referred to as a rectangular bubble

area design. (See Figure C in Section III-l, Improved Stripper Tray Design.)• Check adverse impact on tray efficiency (if any) because of the added waste area.

Downcomer Width And Area

The downcomer should have adequate area to prevent premature column flooding. The downcomer top and bottom width shouldresult in a chord length at least 62.5% of the column diameter for the side downcomer on a one pass tray.As a general rule, a sloped or stepped downcomer should be used if Adi is greater than 12% of As. To ensure good liquiddistribution to the tray below, however, the downcomer outlet area also must be at least 6% of As. This assures that the chordlength is at least 62.5% of the tower diameter for chordal downcomers. For two and four pass trays, the total downcomer outletarea for the side downcomers (or side plus center in the case of four pass) should be 10% and 14% of As, respectively. If thetower diameter exceeds 6 ft (1800 mm) and the liquid rate requires a downcomer area much less than 6% of As, consider the useof a modified arc (segmental) downcomer. (See Section III-K for sizing segmental downcomers.) If a segmental downcomer isused, it must be at least 6 in. (150 mm) wide. (See discussion in Section lII-A on downcomers for more details.)

Outlet Weirs And Downcomer Clearances

Criteria for selecting outlet weir heights and downcomer clearances are given in Table 1. The downcomer clearance is the verticaldistance between the bottom edge of the downcomer and the tray deck. This clearance should be no smaller than 1 in. (25 mm)and is based on avoiding excessive liquid velocity at the tray inlet and to provide an acceptable downcomer seal. Most refineryand chemical plant applications should have a downcomer clearance of 1.5 in. (38 mm) or larger.

Tray Balancing

Even when a new tray design or revamp meets all ExxonMobil criteria, the designer should check to see if the design is wellbalanced. A well balanced tray design will have the jet flood and downcomer flood at approximately the same percentage of theirrespective limits (e.g., 85% jet flood and 85% downcomer flood). This prevents building a potential bottleneck into a tower andpermits the unit to be pushed to its maximum by plant personnel. The designer should run parametric cases in EMoTIP tobalance a design for all potential operating points. Likewise, the designer should try to get all sections of the tower as balanced aspossible (i.e., above the feed vs. below the feed, etc.). Some towers, such as low pressure, low liquid rate fractionators, willalways be controlled by jet flood and/or entrainment rates. It will not be possible to balance these designs without violatinggeometric constraints on downcomer sizing. In this case a balanced design should not be attempted.

DESIGN PRACTICES

Multipass Trays

Multipass trays allow an increase in tower capacity by lowering the tray or downcomer liquid load by splitting the tray liquid intotwo or more paths. Although multipass trays increase tray and downcomer capacity and lower tray pressure drop, they result inshorter flow path lengths. Shorter flow path lengths (smaller than 22 in. [559 mm]) reduce tray efficiency, and if very short, may beinadequate for accommodating tray manways. A flow path length smaller than 16 in. (406 mm) is not wide enough for a traymanway. A minimum flow path length of 18 in. (457 mm) is considered adequate for internal access purposes. Three-pass traysare not available within the EMoTIP design algorithm since their panels are much less symmetrical than two or four-pass trays,which makes it particularly difficult to achieve balanced liquid distribution. Three-pass trays are only available in rating modewithin EMoTIP. Two-pass trays, if an acceptable design can be found, are generally preferred over four or three-pass trays.Anti-jump Baffles must be provided on all center and off-center downcomer(s) of multipass trays if the liquid rate exceeds 4.2gpm/in. of diameter/pass (10 dm3/s/m of diameter/pass). This is to prevent liquid from jumping across (choking) the downcomer,and causing premature flooding (see Section III-A for further information on downcomer choking).

Multipass Tray Balancing

Three- and four-pass trays are more complex than one or two pass trays, in part due to additional tray balancing considerations.For a tray with multiple passes (see Figures 8 and 9), the design which provides the maximum flexibility is the one where the totaltray pressure drop is equal or nearly equal for each pass. However, it is also desirable to maintain about an equal ratio of liquid togas rate per pass for good efficiency. Furthermore, the percent of flood should be roughly in balance to avoid premature floodingby one pass only. Single pass trays obviously meet these criteria since there is only one flow path for the liquid and one for thevapor to travel. The criteria are also met in two pass trays since there is a common chamber at alternate trays that permits equalvapor and liquid flow as well as pressure drop per pass. However, since four pass trays do not have a chamber common to allpasses at a given elevation, special care must be taken to ensure well-balanced operation.To minimize the effects of maldistribution on efficiency, the designer should provide approximately equal bubble areas and equalhole areas for each pass (alternatively, an equal number of valves for valve trays). This will enable each pass to handleapproximately the same vapor loading and have the same dry tray pressure drop. The designer should verify that equal liquidflow is provided to each pass. This can be achieved by specifying a picket fence weir on the center downcomer pass (the B pass,see Figure 9) that reduces the B pass outlet weir effective length to approximately the chordal weir length of the A pass. Analternate method of balancing is to use a different outlet weir height and/or downcomer clearance for each pass to promote orretard the liquid flow on specific pass(es) as needed. However, this is not the preferred method, because fine tolerances andadjustments on downcomer clearance are required and these are difficult to make during installation.Another common tray balancing problem is that the percent of flood on the passes flowing toward the side of the tower (the Apasses for three and four pass [see Figures 8 and 9], may be higher than the other passes. This is because the side downcomerweir length is relatively short, giving a high liquid weir loading and thus raising the calculated percent of flood. If the difference inoverall flood is greater than about 4% between passes, the designer may choose to reduce the relative bubble area (keeping Ao /Ab constant) for the outboard passes to reduce the vapor rate to those passes. The liquid flow should then be re-balanced.Alternatively, the vapor split per pass can be varied by changing the Ao/Ab ratio per pass. This technique is more useful indesigning revamps, where bubble and downcomer areas are already fixed.

The distribution ratio calculated by EMoTIP is the ratio of qv/QL for two passes the pass with the highest divided by the pass withthe lowest gas to liquid ratio. The distribution ratio should be within 8% of unity at design conditions, in order to have the traypasses balanced from a hydraulic and efficiency point-of-view.Another means of balancing the vapor split is to provide the center and off-center downcomers with vapor crossover pipes throughthe downcomers (also known as vapor tunnels), so vapor can flow from one chamber to another. This provides a means of vaporcrossover between adjacent passes and helps equalize the pressure among other passes at the same elevation. The "vaportunnel" area is an input in EMoTIP, which uses the area to calculate the degree of pressure equalization between the chambersafforded by the vapor tunnel or crossover. The equations for liquid, vapor, and pressure distribution are presented in Table 3 forthree and four-pass trays.

Column Access

Entry into the shell of a distillation tower is possible only through manholes. Recommended manhole diameters are in the rangeof 18 to 30 in. (460 to 760 mm). Usually, each manhole serves 10 to 20 trays. For clean and noncorrosive services, eachmanhole may serve more than 20 trays. The manhole diameter affects the number of parts that are used to assemble each trayand other tower internals. Larger manholes are necessary if personnel entering the column need to wear special bulky protectiveequipment. Frequently, tray spacings must be locally increased to be larger than the manhole diameter. Therefore, it is goodpractice to install manholes in the space above the feed trays where the tray spacing is normally lengthened.

Startup Considerations

At very low vapor velocities (such as during startup), sieve trays may dump, with the result that no liquid level is maintained on thetray feeding the reboiler drawoff box. Hence, when thermosyphon reboilers are used on sieve tray towers, special provisions maybe necessary to ensure that the reboiler will have liquid feed during startup. This can be done by either:• Installing a jumpover line from the tower bottoms drawoff line to the reboiler inlet. The jumpover line must have a valve, so

that it can be closed when the reboiler is generating enough vapor to support the liquid on the drawoff tray, or• By providing a chimney tray as the drawoff tray. For the design of chimney trays, drawoffs and other tower internals, see

Section III-H.

PROCESS CONSIDERATIONS

Tray Turndown And Weeping

Turndown is the ratio of the maximum to minimum vapor loadings between which good tray efficiency is maintained. It is limitedby flooding at high vapor and liquid rates and by excessive weeping at low vapor rates. A turndown ratio of between 2/1 and 3/1 isusually achievable with sieve trays.Turndown requirements are dictated by the combination of two effects. The first is operating turndown and the second is theinherent variation in the loading profile over a tower section. Operational turndown should not be overestimated since this couldresult in decreased tray open area. If the loading profile variations are significant and the trays cannot meet the requiredturndown, consider breaking the original section into two (or more) smaller sections. If this reduces the loadings range to anacceptable level, develop a tray design for each of the new smaller sections. If the number of sections becomes too large,however, valve trays should be considered.Weeping is the portion of the liquid flow on a tray that "leaks" or "weeps" downward through the perforations. The remainingportion proceeds to the tray below in a normal fashion via flow over the weir and into the downcomer. The weeping rate can becharacterized by the parameter "fractional weepage," fw, defined as the fraction of the total liquid rate that weeps. That is,

where: Q w = Weep rate, gpm (dm3/s) at conditionsQ L = Total liquid rate, gpm (dm3/s) at conditions

Referring to Figure 1, the vapor rate at which liquid starts to weep through the perforations is called the weep point. As the vaporrate is reduced further, the weeping rate increases and the tray efficiency begins to drop. The region between fw > 0 and fw < 1 iscalled the weeping region. For a well designed tray, this region normally begins at or below 50% of the jet flooding vapor velocity.The weir load affects clear liquid height and thereby tray residence times. Liquid bypassing is another effect that reducesefficiency in the weeping region. This is due to the fact that the liquid that weeps is not fully contacted with vapor on the tray, andthus proceeds to the next lower tray at a different composition than the liquid entering the tray through the downcomer. Thisreduces the apparent efficiency. Weeping at the inlet of a tray is more severe than weeping at the outlet of tray. Weeping at theinlet of a tray, misses the cross flow efficiency boost of two trays (the liquid falls into the outlet side of the tray below). Weeping atthe outlet side of a tray has little effect.As the vapor rate is reduced still further, the point at which all the liquid weeps through the holes defines the dump point (seeFigure 1). Vapor rates at and below this point are said to be in the dumping region (fw = 1).The only practical way to reduce weeping is by reducing the hole area on the tray. The hole area should be reduced until the holevelocity at minimum rates is equal to or exceeds the hole velocity at 20% fractional weepage, or until another hole area restrictionis reached. The final hole area must insure that the hole area to bubble area ratio exceeds 3.5% and that operation in the sprayregime is avoided. Lower hole areas can still be used, however, by blanking a portion of the tray while keeping a 3.5% minimumhole area in the remaining active portion.It should be noted that 20% fractional weepage can be tolerated without significant efficiency loss. However, on drawoff trays amore restrictive limit on weeping may be appropriate (see GP 05-02-01, Par. 10.2). This may dictate the use of less hole areathan that allowed on the adjacent trays. If the weeping rate cannot be reduced to acceptable levels by decreasing hole area, theuse of valve trays should be considered or a chimney tray used. Creating a bottleneck in a tower due to a drawoff tray designshould be avoided; usually tray spacing is increased to help compensate for reduced hole area and/or bubble area. (See SectionIII-H, Figure 16.)

DESIGN PRACTICES

For revamps and some grassroots cases where initial operation rates are less than design, the simplest way to reduce weeping isto decrease the tray's hole area by blanking. If blanking cannot reduce the fractional weepage to 20% (or less) without hittingother hydraulic limitations, then the designer should study:• Whether the predicted efficiency with weeping is still satisfactory, i.e., are there more trays present than needed (run EMoTIP

to confirm the effect of weeping on efficiency and refer to Section III-I), or• Whether economic considerations permit increasing loadings by "over-refluxing" the tower during turndown operations, or• Whether valve trays with their greater flexibility are economically justified.

Tray Efficiency And Heat Transfer

The designer should recognize that efficiency calculations are necessary for each section in a fractionation tower. In addition, thetrays selected to check hydraulics are sometimes not suited for efficiency calculations due to concentration profile reversals orother reasons. The tray efficiency should be calculated by the procedures given in Section III-I. The number of actual traysrequired for a tower or tower section is then calculated by dividing the number of theoretical trays (which are developed during theprocess simulation stage of the design) by the efficiency expressed as a fraction. See Sections lIl-l and III-F respectively for moreinformation on tray efficiency and heat transfer.

Low Liquid Rate Tray Design

When designing a tower to operate at low liquid rates, it may become necessary to design the tray specifically with minimalentrainment in mind. Note that for E-Method entrainment, Eq. (24) should be used to predict entrainment when L is equal to orless than 1.5 gpm/in. of weir/pass (< 3.7 dm3/s/m/pass of weir/pass). Refer to the discussion on Froth to Spray Regime Transitionlater in this section for ways to reduce entrainment. Installing picket fence weirs is one method to reduce entrainment rates andalso avoid spray regime operation. Another option that can reduce entrainment is to use smaller holes.One of the most common ways to reduce entrainment is to increase the hole area. Unfortunately, this increases the rate of liquidweeping. Even with an optimum design, the tray may weep and entrain at the same time (i.e. there is no "operating window" orturndown available). See the discussion in Section III-A on the "operating window" for more background. If sufficient flexibilitycannot be obtained with sieve trays, the designer should consider valve trays or packing. Because of the complex designproblems involved, your FRACTIONATION SPECIALIST should always be consulted.Furthermore, when excessive entrainment occurs at low liquid loading, an insufficient clear liquid height could result in anunsealed downcomer or poor fractionation efficiency due to an inaccurate calculation of liquid residence time on the tray.

High Liquid Rate Tray Design

There are cases where high liquid rates require use of either a large downcomer clearance (over 3 in. [75 mm]) or a deeprecessed inlet box. In past 1133 designs, shaped downcomer lips were also often used in these services. While shapeddowncomer lips may still help reduce head loss under the downcomer and are mandatory for foaming services, because of thenew limit on velocity under the downcomer, they will not be as widely applied. A shaped downcomer lip must not be used wheneither a recessed inlet box or an inlet weir has been specified. This is because the obstruction presented by the vertical face ofthe recessed inlet box, or by the inlet weir, would cause turbulence and defeats the purpose of the shaped downcomer lip. Thedowncomer clearance with a shaped lip should also be set so as not to exceed the Vud limit of 1.3 ft/sec (0.4 m/sec). Formultipass trays requiring a shaped lip, it should be specified for both center, off-center and side passes. The most commonshaped lip radius specified is 1 in. (25 mm), although EMoTIP can handle any lip radius. Radius lips larger than 2 inches (50 mm)are not recommended. 1133 designs did not allow specification of the lip radius, but assumed a fixed 2 in. (50 mm) radius lip inthe calculation of hudL.

Trays With Drawoff Sumps - A drawoff box generally creates waste area (Aw) on the tray and may also obstruct the flow ofvapor from the tray below. This dictates a conservative design approach. The design criteria for such trays are outlined inSection III-H, Tower Internals.

Foaming

Foaming in fractionation and absorption towers can significantly reduce capacity and lead to premature flooding, liquid carryover,and solvent losses. Tray design for foaming services is difficult, but the key is proper downcomer sizing. Include features such asa large downcomer inlet area, large downcomer residence time, a large downcomer clearance with a radius tip, and a high holearea to keep the dry tray pressure drop below 2.25 in. (57 mm) of hot liquid. Larger hole sizes are often recommended to reducethe tendency toward an emulsion flow regime on the deck that small holes promote.

Foam factors are used by the EMoTIP program to account for the foaming tendency of a chemical system. They are applied toboth the jet flood and in several key locations in the downcomer flood and downcomer choke calculations. A foam factor of 1.0

signifies a non-foaming system; 1.2 or greater is a definite foaming system. Foam factors based on hydrocarbon molecular weightare used for heavy hydrocarbon fractionators such as crude and vacuum towers (pipestills), because production well injectionchemicals or corrosion inhibitors often induce foaming in these towers. Foam factors are applied to very low surface tensionsystems close to the critical point to account for the effect of inaccuracy in the prediction of surface tension. Foam factors areoften "best guess" numbers and are normally derived from experience, not measurement. Many "foaming" systems such as gastreating solutions exhibit foaming only under degraded conditions. The foam factors provide enough hardware upside flexibility toaccommodate some solution degradation, but will not prevent flooding in all cases. See Table 2 for the list of foam factorsrecommended for use in EMoTIP based on service and tower section. If the foam is very stable, even a very low downcomer inletvelocity and a large downcomer may not prevent tower flooding. If the designer is confronted with a new chemical system, forwhich a foam factor is not available, a FRACTIONATION SPECIALIST should be consulted regarding appropriate lab or pilot plantscale tests. If the designer expects a chemical system to be a stable foam, then:a) Process changes should be considered to eliminate the source of the foaming (removal of entrained hydrocarbons into

aqueous systems, elimination of suspended solids, etc.)b) Consider using packing and consult your FRACTIONATION SPECIALIST.c) If the foam source can't be eliminated, then an anti-foam agent may be required. This is usually an effective but expensive

solution to the problem since anti-foam must be added continuously.

Vapor Recycling

When the liquid velocity entering the downcomer is greater than the velocity of the bubbles rising through it, vapor recyclingoccurs. The vapor cannot disengage and this results in vapor being swept through the downcomer and recycled onto the traybelow. EMoTIP calculates the volume fraction vapor carryunder based on the liquid volumetric rate. This calculation is based onhigh pressure FRI sieve tray data with non-foaming systems. This vapor recycle is not normally enough to affect the tray capacity,but a good downcomer design should keep the volume fraction vapor carryunder below 0.15 for high pressure towers. SeeSection III-A for more background on vapor recycling.

FoulingFouling is the accumulation of any type of solid deposit on a tower internal device. Fouling on a sieve tray reduces the effectivehole size of the sieve holes and will eventually plug the tray. Fouling results in diminished tower performance (efficiency, capacity,etc.) or even complete inoperability. Larger holes (3/4 to 1 in.; 19 to 25 mm) should be used in sieve trays operating in moderatelyor heavy fouling services. Solid deposits may also accumulate under the downcomer in fouling services and therefore restrict thedowncomer exit flow area. This may cause excessive downcomer backup, premature flooding, and liquid maldistribution to thetray. To avoid blockage in this area due to fouling, a downcomer clearance of at least 1.5 in. (38 mm) should be used. Also,recessed boxes and inlet weirs should not be used in fouling services. The probability and consequences of fouling in the columnmust be fully evaluated. EMoTIP includes a fouling factor (FF) to automatically set the fouling tendency based on historicalexperience with service and location in the tower. Both the tower service and the tower internal location must be specified so thatEMoTIP can correctly set the fouling factor. The fouling factor is currently only used in the design algorithm to set hole size,minimum downcomer clearance, minimum recommended tray spacing, and the use of inlet weirs. Refer to Table 2 for the foulingfactor recommended for each tower service. The table below shows the effect on the tray design algorithm for a given foulingfactor:

FoulingFactor Description

Hole Size, in. (mm)

DCC, in. (mm)

Starting point Other Considerations

0 Ultra-clean 0.375 (9.5) 1.5 (38)

1 Clean 0.5 (13) 1.5 (38)

2ModerateFouling 0.75 (19)

2.0 (51)Do not use float valve design. Consider fixed valves. No Inlet weir.

3HeavyFouling 1 (25)

2.5 (64)Do not use float valve design. Consider large fixed valve. No Inlet weir.

4SevereFouling XXXXXXXX XXXXXXXX Do not use trayed design. Consider sheds or grid.

DESIGN PRACTICES

Corrosion

Corrosion is a process where some materials gradually wear away usually by chemical action or chemical action combined withfluid velocity (erosion/corrosion). The likelihood of corrosion and its potential effect on column internals must be reviewed. Holessmaller than 3/8 in. (9 mm) in diameter on carbon steel trays may rust over during hydrostatic testing and should be avoided.Thinner 410 SS trays should be evaluated as an alternate to thick carbon steel trays (with high Corrosion Allowance) to avoid theequipment reliability issues due to excessive corrosion products forming in wet service towers. Price will typically be about thesame. Always confirm materials selection with a MATERIALS SPECIALIST.

TOWER CHECKLISTTable 7 of Section III-A contains a detailed tower checklist that should be reviewed for all new designs as well as revamps.

CAPACITY/ PERFORMANCE RESTRICTION MECHANISMSThis subsection presents the different mechanisms that restrict column throughput and/or affect tower performance. It alsoprovides the designer with the most important equations and design criteria used in determining the limitations of a particulardesign. Suggestions for improving tray and downcomer designs to meet ExxonMobil design limits are also included in thissubsection. All the major capacity limits described in this section are calculated using EMoTIP and graphically depicted in Figure2A, B, and C for a single tray design at three different pressure levels. This is commonly referred to as a tray performancediagram. Figure 2 shows how the various capacity limits change as a function of vapor to liquid ratio and pressure level. Thereader may want to compare these figures with the generic Figure 22 in Section III-A.

OVERALL CAPACITYThe overall capacity of any fractionating tower is determined from a combination of different vapor and liquid floodingmechanisms. For this reason, an "overall flood" correlation has been developed for cross flow fractionation devices.

Overall Flood

Overall flood is a statistical combination of jet flood, downcomer flood and ultimate capacity flood, and depends primarily on thelimiting flooding mechanism. These different parameters are discussed independently in detail in Section III-A. The overall floodmodel uses the following equation developed to combine jet flood and downcomer flood.

Flood DC- Flood Jet - Flood) DC Flood, (Jetmax )DC (Jet, Flood Overall γ= (Customary or Metric) Eq. (1)

The second term on the right hand side of Eq. (1) is a correction term designed to improve the statistics and therefore theprobability of successful designs. The correction is limited to a maximum of 6% flood. If the jet flood is close to the downcomerflood, the tray design is well balanced and only a small correction is needed. On the other hand, if one or the other flooddominates, a larger correction is necessary. The optimal value of the correction coefficient gamma, obtained from a statisticalstudy, is given below:

Flood. DC JetFlood for 0.12-

Flood; DC JetFlood for 0.17

≥=γ (Customary or Metric) Eq. (1a)

The positive value of gamma for trays limited by jet flood decreases the overall flood, because the jet flood model slightly under-predicts the tray capacity. The negative value for gamma for trays limited by downcomer flood is required because thedowncomer flood model over-predicts the tray capacity. The overall flood model also takes into account the ultimate capacitycheck; the final Overall Flood is then the maximum of Overall Flood (Jet,DC) and the ultimate capacity flood,

Overall Flood = max [Overall Flood (Jet, DC), Ultimate Capacity Flood] (Customary or Metric) Eq. (2)

Designs up to a maximum of 85% Overall Flood by this correlation are acceptable. For services where fractionation is not critical,such as pumparound trays, designs up to a maximum of 90% Overall Flood by this correlation are acceptable.

Probability Of Non-Flooding Operation

EMoTIP reports the probability that a given sieve tray design will not be flooding at a given percent of overall flood. Theprobability model used in EMoTIP is based on almost all known flood runs from the FRI sieve tray database. The probability

model is a function of overall flood only. The only screening performed on the FRI flood runs was to remove runs with a liquid rateover the weir of less than 1.5 gpm/inch (3.7 dm3/s/m) of weir and to also remove runs which had less than 1.25 in. (32 mm) of hotliquid dry tray pressure drop at 85% overall flood. The probability of non-flooding operation of a tower at a given percent of overallflood can be estimated from the table below.

DESIGN % OFOVERALL FLOOD

% PROBABILITY OFNON-FLOODING

DESIGN75 99.680 98.385 94.490 85.795 70.5

100 50.4105 30.3110 14.8115 5.8120 1.8

It is important to note that this table does not predict the probability of successful operation. Much more than avoiding trayhydraulic flood is involved in the successful design and operation of a sieve tray tower. For instance, inlet, reboiler and drawoffinternals must be correctly designed; the foam factor must be correctly estimated; the tray efficiency must be correctly determined;the design basis must be accurate; control systems and instrumentation must be without defect; and such things as fouling ordamage must not have occurred. It is also important to note that in the case of foaming service tray designs, any uncertainty inthe foam factor will reduce the probability of non-flooding for a given overall flood.

VAPOR HANDLING LIMITATIONS

Jet Flooding

Jet flooding is the limitation that most commonly sets the vapor handling capacity for cross-flow trays. Jet flooding occurs whenthe vapor rate is sufficiently high to "jet" or "entrain" liquid from a given tray to the tray above. It is the primary cause of towerflooding for lower pressure towers. The following independent variables are used in the jet flood model: liquid density, vapordensity, tray spacing, free area, bubble area, hole diameter, hole area, vapor rate, liquid rate, outlet weir length, downcomer inletarea, and tower area. Jet flooding is a strong function of tower diameter and tray spacing and a lesser function of the number ofliquid passes used. See Section III-A for more background information.

Jet Flood Equations - The jet flood equation includes a foaming factor term. Refer to Table 2, System Factors for a list offoaming factors for different services. Whereas the old Table 2, "Design Criteria for Specific Towers" in Reference 17 sets theallowable percent of jet flood based on the service, the new jet flood model keeps the allowable jet flood constant and changes aservice factor, known as the foam factor, to achieve the same end. The foam factor is a qualitative measure of the foaminess ofthe system at hand and is based on experience. It is always equal to or greater than 1.0. Following are the equations to calculatepercent jet flood for a single pass.

ff bFCbC

Flood Jet

= (Customary or Metric) Eq. (3)

where: ff = foaming factor ( ≥ 1.0 )

Where the capacity factor is based on the bubble area:

vρLρvρ

bAvqCb (Customary or Metric) Eq. (4)

DESIGN PRACTICES

The capacity factor at jet flood is defined as:

( ) ( ) ( )Lo

f0.50.04

vbF X APC

0.028d0.046exp

ρρρ 0.0795C

−= (Customary) Eq. (5)

The hole area correction term, APC, is given as:

0.25ρ0.3exp 10100*

AA 0.021APC

o (Customary) Eq. (6)

The liquid rate correction term, XL is given as:

L , 1XL = (Customary) Eq. (7)

L , ( )

−2.0

2.04136.0exp v

AX ρ (Customary) Eq. (8)

There is no specific design limit placed on the jet flood value, since it is incorporated in the Overall Flood. However, the calculatedvalue can be used to help determine if the tray design is well balanced and if the tray is jet flood limited or limited by some otherflood mechanism or secondary design parameter.

This jet flood model includes effects for hole size, hole area, downcomer inlet area as a fraction of tower area, and bubble area.These effects were not present in the previous jet flood model based solely on tower free area, liquid rate over the weir, trayspacing and physical properties. Smaller holes will yield more capacity by Eq. (5), for instance about 6% more capacity can beachieved on going from a 0.75 inch (19 mm) hole to a 3/8 inch (9.5 mm) hole. The APC term, Eq. (6), has a coupled effect of holearea and vapor density. A hole area of 10% is vapor density neutral. At a low vapor density, increasing the hole area has a largereffect on jet flood capacity, than at a high vapor density. For instance, at a vapor density of 0.15 lb/ft3 (2.4 kg/m3), a 14% boost injet flood capacity will be obtained by going from an 8% hole area tray to a 14% hole area tray (on bubble area). The same changein hole area will yield a 2.4% increase in jet flood capacity at a vapor density of 1.35 lb/ft3 (21.6 kg/m3). At low vapor density, withlow liquid rates, changing downcomer size has almost no effect on jet flood. At high liquid rates, going from a 6% downcomer toparea to a 25% top downcomer area will increase capacity by 2.5% at 0.15 lb/ft3 (2.4 kg/m3) and 4.8% at 3 lb/ft3 (48 kg/m3) vapordensity.

Ultimate Capacity

Ultimate capacity is the maximum available capacity for vapor flow in a given column diameter with a known liquid rate andphysical properties. Two versions of the ultimate capacity are determined for trays, the Tray Ultimate capacity and the UniversalUltimate capacity. For trays, ultimate capacity usually only limits in hydrocarbon distillation systems above 250 psia (1730 kPa).

Tray Ultimate Capacity - provides an upper bound to the capacity of a cross flow fractionating tower regardless of tray designand tray spacing. It is the highest vapor load a conventional trayed column can handle. Tray ultimate capacity cannot beincreased with hardware modifications that do not affect the free area since it is solely dependent on the vapor load, systemproperties (composition, temperature, and pressure), and the tower free area. Any tray modification that increases free area (suchas sloping the downcomer) will make a small improvement in tray ultimate capacity. In addition, because the packing ultimatecapacity equation includes a liquid rate term, switching from trays to packing or vice-versa will yield a different ultimate capacity.For new column designs, the designer must determine whether the ultimate capacity has been reached. If so, the designer shouldincrease the column diameter of the design. For revamps, a tray with greater free area or packing may provide some relief to a

tower limited by tray ultimate capacity. Also, some non-conventional internals that rely on enhanced deentrainment may be ableto function at high values of ultimate capacity. The equations below are used to calculate the tray ultimate capacity of aconventional tray, cross-flow fractionating tower.

Lult ρρ

β0.65C

+= (Customary) Eq. (9)

where: 5.0

vLρρρ4.1

−=β (Customary and Metric) Eq. (10)

For the metric equation, use a coefficient of 0.396 vs. the 0.65 in Eq. (9).

CCFlood Capacity Ultimate

Where the capacity factor based on the free area is:

vf ρρ

ρ AqC (Customary or Metric) Eq. (12)

Since the overall flood limit is typically 85%, the ultimate capacity flood limit is also 85%. It is important to note that the trayultimate capacity is one of the correlations for sieve trays with a high degree of uncertainty. Of the 894 flood runs in the FRIdatabase only 27 are limited by tray ultimate capacity (i.e. the ultimate capacity sets the overall flood). Many of these runs alsohave high values of either jet or downcomer flood.

Universal Ultimate Capacity - is another way to view ultimate capacity, and considers the entire tower area rather than just thefree area; it appears to be a better indicator of the true ultimate capacity of a given tower shell. The universal ultimate capacitymay be applied to both trayed and packed towers. Universal ultimate capacity is independent of tray design and tray spacing ortype of packing. It is dependent only on the system properties (composition, temperature, and pressure) and on the tower crosssectional area. The system properties determine a drop size, which places a limit on achievable capacity independent ofhardware design. The universal ultimate capacity uses the tower cross sectional (superficial) area instead of the free area that isused in the existing tray ultimate capacity correlation; this move toward using superficial area is consistent with FRI's newestultimate capacity correlation. The correlation also includes the effect that increasing the liquid rate has on decreasing the vaporcapacity. The universal ultimate capacity correlation also considers the Reynolds number dependency on the drag coefficient.

The critical Weber number forms the basis for this calculation. FRI data suggests that:

7.0Wec = (Customary or Metric) Eq. (13)

Using a lower Wec will predict ultimate capacity limitations for a greater percentage of runs, i.e. is more conservative from a designperspective. The critical Weber number and the critical Reynolds number are defined by:

( )( )0.00220462σ2

uρ/12DWe

c ⋅⋅⋅⋅

= (Customary) Eq. (14)

( )( )0.00067197µ

/12DuρRe

ptvc ⋅

⋅⋅= (Customary) Eq. (15)

DESIGN PRACTICES

Where the drop size can be calculated from:

12∆ρgρc

Dp ⋅

⋅⋅

The drag coefficent depends on whether the drop size yields an Rec which is in the Stokes Law Intermediate region or Newton'slaw region of applicability:

2Re Re24C

500Re2 Re

500Re 44.0C

(Customary or Metric) Eq. (17)

The following equation calculates the terminal velocity of a liquid droplet from the superficial area, the vapor volumetric rate, thevapor and liquid densities, and the liquid volumetric rate.

⋅⋅=

Lβ1ββ1A

ρ3600w`1000u (Customary) Eq. (18)

Equations (14) through (18) represent five equations in five unknowns: ut, Dp, w`v, Rec, and cD and can be solved simultaneouslyfor the unknowns.The limiting capacity factor for the universal ultimate capacity is then:

vuniv ρρ

ρ3600ρ A w`1000C

Percent Universal Ultimate Capacity is the ratio of the capacity factor based on superficial tower area to the capacity factor atultimate capacity:

vs ρρ

CCCapacity Ultimate Universaluniv

If the Universal Ultimate Capacity is greater than 85%, a larger diameter tower should be designed.

Spray Regime And Entrainment

Spray regime and entrainment are both secondary design parameters that are primarily vapor handling limitations. Spray regimeis a transition from froth to a spray of discrete droplets. It is also referred to as "blowing" when it is extreme. In such cases, thespray appears to be suspended above the deck, a condition known as "blowing dry". Entrainment is the lifting of tray deck liquidto the next tray above. It increases rapidly with increasing vapor rate as the flood point is reached, but can also be high as apercentage of tray liquid rate, even at low values of jet flood, when the tray liquid load is low. Refer to Section III-A for additionaldiscussion.

Froth To Spray Regime Transition - Spray regime operation occurs primarily at high vapor velocities and low liquid rates. Suchconditions are likely to occur in distillation towers operating below 50 psia (345 kPa), water wash towers, and atmospheric pipestill

wash zones. In the spray regime, the liquid becomes suspended as a dispersed phase above the tray deck, interphase contactbecomes poor, and the tower fractionation efficiency deteriorates. See Figure 16, Section III-A.R&D studies have shown that the transition from froth to spray is primarily an inertial phenomenon related to the ability of thevapor jet to penetrate the liquid on the tray. The resulting correlation, which gives the vapor load per unit tower bubble area atwhich the transition from froth to spray occurs, is:

[ ]2.0

Eq. (22)

Where: 1c = 0.214 (Customary)

1c = 0.04 (Metric)

Note the dependence of the equation on tray spacing, hole size, and liquid rate per inch of weir. For trays with more than onepass, all passes should be checked even though the pass leading to the center downcomer will usually limit.

For definitions of the terms and information on units to be used in this equation, refer to the NOMENCLATURE section. Theappropriate fraction of the transition point to use for design calculations can be found from the table below. The maximumallowable percent of Spray/Froth transition velocity has been increased by 10% from the values given in Reference 17. This isdue to tighter limits on entrainment (now 10% maximum vs. previously 20% maximum) and a more accurate and robust jet floodcorrelation at low liquid rates.

MAXIMUM ALLOWABLE PERCENT OF THE SPRAY/FROTH VELOCITY

VAPOR DENSITYlb/ft3 (kg/m3)

L ≤≤≤≤ 1.5 gpm/in.(≤≤≤≤ 3.7 dm3/s/m)

L > 1.5 gpm/in.(> 3.7 dm3/s/m)

ρv ≤ 0.08 (≤ 1.28) 66 93.5

0.08 ≤ ρv < 0.6(1.28 ≤ ρv < 9.6)

60.5 (1.0 + 0.9 ρv)60.5 (1.0 + 0.057 ρv) 93.5

ρv ≥ 0.6 (≥ 9.6) 110 110

Alternatives are available to the designer to avoid operating in the spray regime. Note that increasing the weir height will not helpsolve this problem. These alternatives are presented in the order in which they should be considered:

NEW DESIGNS REVAMPS

• Increase the hole area • Increase the hole area

• Decrease the hole diameter • Decrease the hole diameter

• Install picket fence weirs* • Install picket fence weirs*

• Increase the tray spacing • Consider packing

• Increase the bubble area

• Use packing

*For more details contact your FRACTIONATION SPECIALIST.

Mini-valves (either fixed or moveable) should also be considered as a spray regime remedy.

Intertray Entrainment - The quantity of entrainment generated is dependent on vapor rate, liquid rate, and certain hardwareparameters. There are two methods for predicting entrainment used by the ExxonMobil Tower Internals Program, the E-Methodand the M-Method. Neither one of these methods is very accurate due to the inherent problems in measuring entrainment rateand the exponential nature of entrainment as a function of vapor rate. However, if the calculated entrainment values from both

DESIGN PRACTICES

methods lie within the acceptable limit of 10%, it is unlikely that the column will experience entrainment problems. Descriptions ofthe E-Method and M-Method correlations are below.

E-Method Entrainment - Entrainment is based on data from Fractionation Reasearch Inc. (FRI) and ExxonMobil test programs.This correlation takes into account the effect of system physical properties and tray hardware parameters on entrainment rates.Nevertheless, if the fractional entrainment fe, (i.e., the entrainment rate divided by the design liquid rate) exceeds 10%, the holearea should be increased and the fractional entrainment rate recalculated. Use this equation if the volumetric liquid rate is greaterthan 1.5 gpm/in. of weir/pass (is greater than 3.7 dm3/s/m of weir/pass), otherwise use Eq. (24).

fe = Kφ KL Kσ Kε

bw1000

A (Customary) Eq. (23)

fe = Kφ KL Kσ Kε

(Metric) Eq. (23M)

Where:

fe = Fractional entrainment, dimensionless. For 2 pass trays, calculate for both center and side passes.

Kφ = Tray geometry factor (see Figure 3).

KL = Liquid rate/tray spacing factor (see Figure 4A or 4B). For two pass trays, determine KL for each pass.

Kσ = System properties factor (see Figure 5).

Kε = Vapor energy dissipation factor (see Figure 6A or 6B).

Ab = Tray bubble area ft2, (m2). For two pass trays, calculate for both inboard and outboard passes.

wL = Liquid mass flow rate, k Ib/hr (kg/s).

The equation below should be used to calculate fractional entrainment (fe) when the liquid rate is less than or equal to 1.5 gpm/in.of weir/pass ( ≤ 3.7 dm3/s/m of weir/pass). For term definitions, refer to NOMENCLATURE.

[ ] [ ]

0.550.59

o0.54wo

hL0.8042

lA3.96

[ ] [ ] [ ]

σµ0.55

o0.54wo

b0.45o

KHAAhL

lA 1083d

AV 324

= (Metric) Eq. (24M)

For the inboard pass on tow pass trays, substitue Ic* for Ic.

when: 1Cotherwise;ρ

0.08C0.08,ρ 4v

4v ==> (Customary)

where: fe = Fractional entrainment, dimensionless

n = [ ] )6.6n0.12;vρif(vρ9.4 1/6 => (Customary)

Note: If hwo > 2 in. (> 50 mm); set hwo = 2 in. (50 mm)If hwo < 1 in. (< 25 mm); set hwo = 1 in. (25 mm)

M-Method Entrainment - Entrainment depends on the vapor rate at flooding conditions and is thus dependent on the jet floodmodel.

( )[ ]

vρvρlρ

olLQ3123*bFCbC

vlvbbF 10*ρρρ ACwE (Customary) Eq. (25)

LIQUID HANDLING LIMITATIONSLiquid flows across the tray and is contacted by the ascending vapor. At the downstream end of the tray, the liquid enters adowncomer, which carries it to the tray below where the contacting process is repeated. The contacting area must be largeenough to handle the required liquid and vapor rates while promoting the desired mass transfer. Likewise, the downcomer mustbe large enough to handle the froth from the tray deck and clarify this froth. Premature tower flooding can occur as a result ofeither inadequate downcomer area or depth.

Downcomer capacity has been an active area of research at FRI and ExxonMobil since the early 1980's. Models of downcomercapacity have been developed which not only predict the downcomer froth density and froth height but also accurately predictwhen the downcomer floods. Earlier design procedures for downcomers (Ref. 17 for example) relied on satisfying a series ofdesign constraints, instead of actually determining the downcomer flood point. More recent models of downcomer flood, such asthe one included in EMoTIP, provide an accurate means to predict the liquid handling capacity of the tray and therefore improvethe design.

Downcomer Flood

Percent downcomer flood is the criterion that determines how close a tower is to flooding as a result of excessive froth height inthe downcomer. Percent downcomer flood represents the ratio of the actual vapor and liquid rates to the rates that would result in100% downcomer froth backup. Downcomer flood solves the following equation for x. (hd and Ψ are shown as functions of x inthis equation.)

)x()x(hhH d

woB Ψ=+ (Customary or Metric) Eq. (26)

where: x = Multiplier on Liquid and Vapor rates at which tray is rated, dimensionless HB = Tray spacing below the deck, in. (mm)

Then, downcomer flood for the loads at which the tray is rated is given by:

x1Flood DC = (Customary or Metric) Eq. (27)

Using x to scale both vapor and liquid rates (keeping a constant ratio of vapor to liquid), is the appropriate way to handle adowncomer flood calculation for most towers. Even for absorbers or strippers where gas or liquid rates alone are changing, it isusually sufficiently accurate and will be slightly conservative.

DESIGN PRACTICES

The EMoTIP downcomer flood model includes a tray hydraulic model to calculate the froth density on the tray and therefore theamount of froth that the downcomer must handle. For downcomer limited trays, reducing the bubble area on the tray increasesthe frothing action and will reduce the downcomer capacity, i.e. increase the downcomer flood. Downcomer flood is also stronglyaffected by the tray pressure drop through the conventional downcomer filling. This increases the numerator in Eq. (26).Downcomer filling components are described next.

Downcomer Filling (hd) - Downcomer filling is defined as the clear liquid height (i.e. collapsed froth height) in the downcomer. Itis composed of the inlet head (hi) on the tray, the tray pressure drop (ht) across the tray immediately upstream of the downcomerbeing considered, the head loss under the downcomer (hud), and the frictional head loss due to two-phase flow through thedowncomer (hdc). See Figure 19 in Section III-A.

ldcudtid ρρ

ρh h h h h−

∗+++= (Customary or Metric) Eq. (28)

Inlet Head (hi) - The inlet head is equal to the clear liquid height (hc) on a tray if there is no inlet weir present. If an inlet weir ispresent, downcomer filling will increase due to the weir height, the crest over the weir, and added pressure drop of the liquidflowing between the downcomer apron and the inlet weir.

ci h h = (tray without inlet weir) (Customary or Metric) Eq. (29)

( ) wi3/2

iLi hl/Q*48.0h += (with inlet weir) (Customary) Eq. (30)

Tray Clear Liquid Height (hc) - Clear liquid height is the height of liquid on a tray expressed in inches (mm) of hot liquid. Theclear liquid height is a function of the liquid rate, outlet weir height, hole pitch, % hole area, and bubble area. The clear liquidheight must be high enough to provide sufficient contact time between the liquid and the vapor for mass transfer to occur.Excessive clear liquid heights should be avoided, because they increase the tray's pressure drop, increase downcomer filling, andmay cause premature weeping. If the tower is heavily liquid loaded and hc is too high, consider increasing the number of liquidpasses to reduce the liquid rate per length of weir for each pass. The clear liquid height used in the downcomer flood model is theHofhuis equation, which avoids a trial and error calculation of clear liquid height and was found to give a better fit to thedowncomer flood data by FRI.

25.025.0

c bP*0785.2h Φ

Ω= (Customary) Eq. (31)

where: 2h2h 3

=Ω , weir height term

P = Calculated Hole pitch, in.; given by: 866.0A 4

Ad P o

o∗∗∗∗π=

oAlb = , weir length term

48.7601*

∗=Φ , Lockhart-Martinelli Flow Parameter

Tray Pressure Drop (ht) - The tray pressure drop (ht) is composed of the dry tray pressure drop (hed) and the clear liquid height(h'c). Each of the pressure drops (or heads) is expressed in inches (mm) of hot clear liquid. Tray pressure drop is critical in traydesign since it is one of the major components of downcomer filling and therefore downcomer flood.

cedt hhh ′+= (Customary or Metric) Eq. (32)

Note that the clear liquid height term (h'c) used in the tray pressure drop calculation is calculated from the FRI method found inTopical Report 88 and therefore differs in value from the clear liquid height term (hc) used for the inlet head in the downcomercalculations. Both terms represent the collapsed froth clear liquid height on a tray. Using the FRI TR 88 method results in the traypressure drop calculation results in better agreement with tray pressure drop data. The FRI method for pressure drop calculatesthe clear liquid height from:

( ) wo63/2

oLwc hSl/Q*F4.0h +=′ (Customary) Eq. (33)

The dry tray pressure drop is calculated from:

v2ov5ed ρ

ρ/AqSφh ∗∗∗= (Customary) Eq. (34)

where:

+′′⋅−

=338.5

eφ , =φ Liquid head correction for dry tray pressure drop

Fw = Weir constriction factor, f( lo/Dt )

S5 , S6 and S9 = Functions of lo/Dt , Ao/Ab , t, do

Head Loss Under the Downcomer (hud) - The head loss under the downcomer uses a correlation of the froth clarification actionof the downcomer to correct the standard clear liquid head loss to an aeration corrected head loss.

Lud ρ

ζ1hudh (Customary or Metric) Eq. (35)

LoeL lc

where ζ = vapor fraction of aerated liquid flowing under downcomer

mρ = Mixture density under downcomer, lb/ft3 (kg/m3)

coe = Downcomer exit coefficient based on the radius of shaped lip dc:0.06 for sharp edge0.036 for 1 in. (25 mm) radius0.020 for 2 in. (51 mm) radius

Head Loss Due to Two-Phase Flow Through the Downcomer (hdc) - This is a new term not found in the conventionaldowncomer filling clear liquid height.

dcDdc dP Kh ∗= (Customary or Metric) Eq. (37)

KD is a correction factor that accounts for the fact that frictional pressure drop will occur mainly in the entrance region of thedowncomer. It is a function of the hardware dimensions (Adi/As), (Adi/Ado) and HB. dPdc is the two phase pressure drop indowncomer, and is a complex function of the downcomer entrance velocity, the tray deck froth density, the tray deck interfacial

DESIGN PRACTICES

area, the liquid physical properties and the total available downcomer height. The dependence of this downcomer two-phasepressure drop on liquid viscosity is conservative and therefore a liquid viscosity cut off of 1 cP is implemented in EMoTIP.Reference 13 contains more background on the validity of the EMoTIP correlation at high liquid viscosity.

The frictional head loss term can be a significant factor in the calculation of downcomer flood. High values can be expected if thetray froth density is low. The term was developed by regarding the two-phase bubbly flow in a downcomer as liquid flowingthrough a packed bed of bubbles. Downcomer flooding will occur if this frictional head loss due to two-phase flow through thedowncomer is excessive. This is often the case if the choke term in EMoTIP is excessive. This model is considered accurate forhigh-pressure systems but may be conservative at medium and low pressures. For details on the development of this model, thereader is referred to Reference 15, FRI Topical Report (TR) 101. The EMoTIP model is a modified version of the model in TR101. A more accurate method of accounting for average downcomer area was developed for the EMoTIP model and a correctionfactor developed based on fitting the model to downcomer flood data using convergence on a constant vapor/liquid ratio basis.(See Reference 10.) Reference 16 contains all the equations currently in the EMoTIP downcomer flood model and can be usedfor hand checks of EMoTIP results.

Downcomer Froth Backup - Percent downcomer froth backup is the froth height in the downcomer divided by the distance fromthe bottom of the downcomer to the top of the outlet weir.

100 )wohH(

dhBackup Froth DC % ∗

Βψ (Customary or Metric) Eq. (38)

There is no separate design criteria for % DC Froth Backup. All downcomer flood effects are captured through the downcomerflood model, combined with downcomer choking and the secondary parameter: Square Root of (% DC Froth Backup x % DCChoke). (See below.) The downcomer froth density used in the downcomer flood and froth backup calculations is divided by thefoam factor (ff).

SECONDARY DESIGN PARAMETERSSecondary design parameters are constraints on tray design that are important, but not critical. Some of these parameters helpkeep the tray design within acceptable bounds where the capacity correlations are valid. Others pick up specific issues that mayoccur from time to time with tray designs. Still others are thought to have a role in tray performance from the physics of fluid flowon the tray, but for one reason or another (such as confounding with another variable) do not have enough data backup to provethat they do not affect tray design. In this last case, the chosen criteria reflect value that are known historically to be acceptable.In general, all grass root tray designs should meet all secondary parameter criteria. In revamp situations, a FRACTIONATIONSPECIALIST should be consulted about whether a given secondary criterion can be violated.

Liquid Entrainment - The design criteria for liquid entrainment is 10% of the liquid flowing on the tray by either the M-method orthe E-method. This is a reduction from the earlier 1133 program limit of 20%, but an increase from the MoTIP limit of 5%. Lowerpressure towers that tend to be limited by jet flood can have stable operation with some level of entrainment. This is often referredto as a state of "incipient flood". As the entrainment rate increases above 10%, tray efficiency will suffer, then when entrainmentlevels are high enough, a full hydraulic flood will result. Because predicting absolute values of entrainment is difficult,conservative entrainment limits are recommended for design. Entrainment effects are included in the M-method efficiencycalculation, but not in the E-method efficiency. Entrainment effects (additional liquid due to entrainment) are not included in eitherthe jet flood or the downcomer flood calculation. E-method entrainment reduction for new or revamp designs can be achieved bymethods similar to those to control spray regime operation. M-method entrainment reduction can be achieved by reducing the jetflood.

Downcomer Choking Due to Velocity - The downcomer choking criteria is based on an ultimate-capacity like analysis of bubbleseparation in liquid. The allowable downcomer entrance velocity limit is given by:

di(Ult)ff1

ρρσ

β1βV VL ∗

−⋅∗

+= (Customary) Eq. (39)

The percent of downcomer choke is then given by:

100 %)(

∗=Ultdi

VChokeDC (Customary or Metric) Eq. (40)

The previous limit for this parameter was 75% in MoTIP. The limit has been increased to 100% in EMoTIP, because statisticalanalysis of this variable has shown it is not a very significant predictor of flood above and beyond the basic downcomer floodmodel itself. However, high values of the DC choke parameter do usually indicate that the downcomer mouth area is restrictiveand extra downcomer capacity can be obtained by increasing the downcomer mouth area. The foam factor also appears in thisequation squared. In MoTIP, the foam factor was to the first power in this equation. With an allowable value of 100%, squaringthe foam factor keeps entrance velocities at about the same absolute value as allowed by MoTIP with a design limit of 75%.These entrance velocities are similar to those used in the 1133 program. Tray design for foaming systems require largedowncomers and a large downcomer mouth area to provide sufficient residence time for foam collapse in the downcomer.

Square Root of (% DC Froth Backup x % DC Choke) - A better predictor of downcomer flooding than the choke parameteralone is the square root of the product of choke and downcomer froth backup. The criterion for this parameter is to keep it below70%.

( )ChokeDCBackupFrothDCChokeDCBackupDCSQRT % %) ( ∗=∗ (Customary or Metric) Eq. (41)

Liquid Rate per Unit Length Weir - To keep within the bounds of the data used to develop the capacity correlations, the liquidrate per unit length of weir should be within the criteria of 1.5 to 17.5 gpm/inch of weir (3.7 to 43.5 dm3/s/m). Picket fence weirsmay be used to increase the liquid rate should it be less than 1.5 gpm/inch (3.7 dm3/s/m). Likewise, swept back weirs or modifiedarc downcomers may be used to decrease the liquid rate per unit length on side downcomers. Increasing the number of passesshould be evaluated provided sufficient diameter exists. Multi-downcomer trays such as the HiFi tray or the ECMD tray are alsooptions for high liquid rate conditions.

Froth/Spray Transition - Statistical analysis has shown that the 1133 program limits on froth to spray transition can be safelyincreased by 10%. This increase is already reflected in the table under Froth to Spray Regime Transition earlier in this sectionand the allowable criteria are also printed out by EMoTIP.

Universal Ultimate Capacity - This new correlation is still in the testing stage but should be a good indicator of whether a tower istoo small for any internal, except perhaps non-conventional devices that use centrifugal or impaction deentrainment. If the limit of85% is reached, an increased diameter is required.

Dry Tray Pressure Drop - To keep within the bounds of data used for both the capacity and efficiency correlation development, itis recommended to maintain the dry tray pressure drop in the range of 1.25 to 5.5 inches (32 to 140 mm) of hot liquid. Maintaininga dry tray pressure drop is an old rule of thumb for sieve tray design to maintain a stable tray frothing action. The research on theEMoTIP probability of flooding provided support for this old rule of thumb, in that trays with low dry tray pressure drop at a fixed85% of overall flood, exhibit markedly increased scatter in the flood point, either high or low. EMoTIP performs this test and willnot print out a probability of non-flooding design for the tray if a minimum dry tray pressure drop of 1.25 inches of hot liquid (32mm) is not maintained at 85% of overall flood.

Velocity Under the Downcomer (Vud ) - Excessive velocity in the inlet area of the tray can result in channeling across the tray,which could reduce tray efficiency. It can also cause a blocking effect of the holes on the inlet side of the tray due to momentum,and result in excessive froth heights on the outlet side of the tray and premature flooding (commonly referred to as a rooster-taileffect or outlet side flood). This latter effect tends to be worse with low dry tray pressure drops, low tray spacing, and longer flowpath lengths. The current criterion in EMoTIP is 1.3 ft/sec (0.4 m/sec), which is an increase from the 1.1 ft/sec (0.33 m/sec) limit inMoTIP. There was no limit on this parameter in 1133 designs. With tighter downcomer sealing criteria, and a recommended hudLof up to 1.5 in. (38 mm), many 1133 designs will have high Vud. The downcomer clearance should be increased if this limit is notreached. Recessed inlet pans or inlet weirs may also be used to prevent a high velocity under the downcomer from adverselyaffecting tray action. However, inlet weirs will cause increased downcomer backup. Both recessed inlet pans or inlet weirs shouldbe avoided in situations of potential liquid born foulant, as they increase the plugging potential of the tray. In those cases, a smallbreaker bar, 3/8 to 1/2 inch (9.5 to 13 mm) high may be used in place of an inlet weir to help alleviate the effects of high velocityunder the downcomer. Please consult with a FRACTIONATION SPECIALIST if this criteria can not be met during revamps or re-ratings.

DESIGN PRACTICES

Vapor Fraction Under the Downcomer (Vapor Recycle) - There is no limit on this parameter in EMoTIP, but downcomer designshould attempt to minimize it. In high pressure and heavily liquid loaded towers it may not be possible to reduce the value below15-18%. This number is the vapor as a fraction of the total volumetric flow in the downcomer, so as a percentage of the totalvapor flow in the tower it is small. Therefore, it usually has no appreciable effect on jet flood or tray efficiency and is not includedin those calculations.

Downcomer Sealing - To prevent some of the vapor from bypassing a tray by traveling up the downcomer, the downcomershould be sealed at design rates by the liquid on the tray below. Therefore, it is necessary to check the sum of the clear liquidheight at the inlet to the tray (hi) and the head loss under the downcomer (hud) at design liquid rates. This sum plus 1/2 in.(13 mm) must be at least equal to the downcomer clearance. Downcomer seal is also calculated at minimum rates, if a turndownrate is defined. Larger degrees of unseal are allowable at turndown. If a seal is not obtained, consider:• Increasing the outlet weir height.• Reducing the clearance to 1.5 in. (38 mm) provided the downcomer filling is not exceeded at design rates and the velocity

under the downcomer limit is not exceeded. A downcomer clearance of 1 in. (25 mm) is acceptable in very clean services.• Adding an inlet weir.• Using a recessed inlet box.Note that designing near the high end of the head loss range may unnecessarily increase downcomer filling if a higher clearancewill still seal the downcomer. Therefore, there is no justification for setting downcomer clearance any lower than that required fordowncomer sealing.Downcomer unsealing up to 1.0 1.5 in. (25 to 38 mm) at turndown in high liquid rate services are acceptable or in serviceswhere there is a need to minimize downcomer fouling. Downcomer unsealing at turndown does not seem to adversely effect trayefficiency according to available FRI data.

Liquid Weeping at Conditions - Weeping as a percentage of the total liquid rate on a tray will generally have limited effect ontray efficiency if weeping is less than 20%. The estimated weeping is calculated both at the design conditions and the turndownconditions (if they differ from the design). It should be maintained at less than 20%. Both the E-Method and M-Method efficiencycalculations have corrections for weeping so an estimate of the effect of weeping on tray efficiency can be made. Hole area hasthe largest effect on weeping and should be reduced to reduce the weeping if it is excessive. If the turndown requirements for thetower cannot be met with a sieve tray, a moveable valve tray should be evaluated. The EMoTIP weeping correlation is a variationof the FRI TR 119 model, and is the same model that was used in MoTIP. The basic equations for weep point and weeping rateare given below, but are only suitable for computer calculation, because the clear liquid height must be determinedsimultaneously. The clear liquid for these equations is the public domain version of hExxon 1133 clear liquid height (referred to asthe Colwell model in the literature). Details on this method for weeping may be found in FRI TR 119 and Reference 12.

( ) 0.144cwpL

bwp hρt

dρρρ

14.5V ′′⋅∗

−∗

∗=−

(Customary) Eq. (42)

where: =bwpV Vapor velocity through the bubble area at the weep point, ft/s

=t Tray thickness, inch=′′cwph FRI TR 119 clear liquid height at the weep point, inch

The weeping rate is determined from the following equation:

t0d0.043

VV1eα1.0994.6Q ∗

′′−′′

⋅∗

−∗∗∗=

⋅⋅− (Customary) Eq. (43)

where: =wQ Weep rate, gal/min

α = Froth density on tray, dimensionless (Using FRI TR 119 procedure)=′′ch FRI TR 119 clear liquid height at the condition where Qw is calculated, inch

=′′th FRI TR 119 Tray pressure drop, inch

uL = Liquid velocity across the tray deck, based on liquid not weeping, ft/s=bV Vapor velocity through the bubble area at the condition where Qw is calculated, ft/s

The only difference between this procedure and the FRI TR119 procedure is the insertion of the 1.09 factor in the weep rateequation. This factor improves the weep rate fit in the mid- and high-range weeping rate, which is the industrial area of concernbecause the effect on tray efficiency is more pronounced in this area.

SIEVE TRAY DESIGN PROCEDUREAt the basic design stage, the following parameters should already be known: vapor and liquid flow rates, operating conditions,and the type of tray to be used. For new tower designs and for rating existing towers, the ExxonMobil Tower Internals Program(EMoTIP) should be used. A discussion of the ExxonMobil Tower Internals Program design algorithm is included in this section.The following are fundamental factors that influence tray performance and therefore, the desired tray layout.The first step in tray design is to obtain the vapor and liquid loadings and their respective physical properties. This information isnormally calculated as part of the heat and material balance(s) for the tower and is usually obtained from a computer programsuch as PRO/II or PROVISION. Traditionally, designers have assumed 50% of the design loadings when specifying minimumliquid and vapor loadings. However, this assumption can have a significant impact on the tray design and operation, and thereforeproper minimum loadings for a given design should be used. Tray spacing and the presence of drawoff boxes, feed zones,transitions, and other internals can have an impact on which tray is capacity limiting a given tower. Designers should verify byinspection (not auto-selection) the minimum and maximum loaded trays in a section to be used for tray design or rating. This isdone by determining the vapor load for the trays in a section (using data from the tray loading summary in the PRO/II output) andselecting the trays with the largest and smallest vapor load. Vapor loadings are to the tray in question; liquid loadings are from thetray in question since these are nearly always the maximum values. However, the designer should be aware that loadings canincrease significantly across a given theoretical tray. If this is the case and the overall efficiency is less than about 70%, the vaporloading to the upper actual tray may be higher. The designer must then prorate loadings between the loadings to and from thetheoretical tray in question. In the case of bottoms and sidestream strippers for pipestills, guidelines are presented in Section lIl-l,Tray Efficiency, for 4 and 6 tray strippers.

Once the vapor and liquid loadings, stream properties, and the turndown loadings have been established, the type of internalsmust be correctly chosen before the ExxonMobil Tower Internals Program (EMoTIP) can be used for design.

EMOTIP DESIGN ALGORITHMEMoTIP is capable of designing one, two, and four pass trays with either straight or sloped downcomers. It does not design threepass trays. EMoTIP also does not include seal pans, sweptback weirs, and modified arc downcomers in any of its designs. Theprogram allows the user to select any combination of the following options simultaneously: shaped lip option, inlet weir option, andpicket fence weir option. Selecting these additional options allows EMoTIP to design with these hardware features if they aredeemed applicable to the specific design by the program.

EMoTIP uses an overall objective function (OBJoverall) to determine the best design from a matrix of achievable designs, eachdetermined from a second objective function (OBJ). It is able to search the entire design space, evaluating every possiblecombination of standard tray geometry. A flowchart of the search algorithm is shown in Figure 10. Since the user cannotreproduce this method by hand, this program feature proves to be very valuable and represents a distinct advantage overprevious tray design programs.

The optimization matrix finds the minimum tower diameter, Dopt, that will yield a valid tray design for each combination of trayspacing and number of passes. In the default condition, there are 21 possible combinations of tray spacing and number ofpasses. An optimum tower diameter is determined for each of these 21 cells in the matrix.

The "design optimization matrix" is shown below. EMoTIP provides this matrix at the end of the design option debug file (*.DBG).Inside each cell of the matrix is the minimum diameter needed to meet all of the designer's criteria. For example, Dopt (TSmin,4)= 15 ft (4600 mm) means that a 15 ft (4600 mm) diameter tower is needed if a 4-pass tower with the minimum tray spacingallowed by the designer is desired. Typically tray spacing is evaluated from 18 to 36 inches (450 to 900 mm) in 3 inch (75 mm)

DESIGN PRACTICES

increments, resulting in seven rows in this matrix. However, the starting value for TSmin may be reset automatically dependingon the fouling or foaming factors, or by user override of the default TSmin.

Number of Passes

TSmin Dopt (TSmin,1) Dopt (TSmin,2) Dopt (TSmin,4)

TSmin+TSstep Dopt (TSmin+TSstep,1) Dopt (TSmin+TSstep,2) Dopt (TSmin+TSstep,4)

... ... ... ...

TSmax-TSstep Dopt (TSmax-TSstep,1) Dopt (TSmax-TSstep,2) Dopt (TSmax-TSstep,4)

TSmax Dopt (TSmax,1) Dopt (TSmax,2) Dopt (TSmax,4)

EMoTIP allows overriding all maximum, minimum, and step sizes for all geometry in design mode. The default design algorithmvalues are given in Table 4A (customary) and Table 4B (metric). For each of the twenty one different combinations or cells in thematrix, the program finds the optimum design based on the objective function OBJ. The objective function OBJ determines theoptimum design for each cell within the matrix given its corresponding tray spacing, number of passes and the minimum towerdiameter, Dopt. The objective function finds the optimum design by minimizing the following equation:

( ) ( )∏×=i

i ionConsiderat DesignfFlood OverallOBJ (Customary or Metric) Eq. (44)

OBJ is primarily controlled by percent overall flood, since designing at a low percent overall flood is the best way to ensuresuccessful tower operation. The fi(Design Consideration) represent debits applied to OBJ when a design is close to a secondarydesign limit. There are 16 such debits:

Design Debits: fi(Design Consideration)f(DTPD) - Dry tray pressure drop (see Figure 11)

f(GPMin) - Liquid load per unit weir (see Figure 12)

f(FS) - Froth/spray regime (see Figure 13)

f(Choke) - Downcomer entrance choking(see Figure 14)

f(Ent) - Entrainment(see Figure 15)

f(FPL) - Flow path length(see Figure 16)

f(Weep) - Weeping %(see Figure 17)

f(Weep2) - Weeping less than 20% at Turndown Condition(see Figure 18)

f(Seal) - Design condition downcomer seal(see Figure 19)

f(Seal2) - Turndown condition downcomer seal(see Figure 20)

f(PicketWeir) - Application of picket fence weir = 1.1 debitf(Carryunder) - Vapor carryunder through the downcomer (See Figure 21)

f(VaporChannel) - Vapor channeling (valve trays only) f(ValveWeep) - Valve tray weeping (valve trays only)f(DCShape) - Sloped side downcomer = 1.03 debit f(InletWeir) - Application of inlet weir = 1.125 debit

The 21 cells which each represent an optimum design for a given combination of tray spacing and number of passes are thenevaluated with a second objective function, OBJoverall.

The dependence of OBJoverall on tray spacing and diameter represents a realistic measure of the relative cost of tower height todiameter. However, the optimization matrix should be consulted before a final design is chosen, so the designer can include morerealistic, application-specific information (such as space available for tower footprint, possible step changes in cost for largediameter or very tall towers, etc) in the determination of which design is the best.

( ) ( ) ( )Flood OverallfNfHDOBJOBJ p802.0066.1

toverall ××××= (Customary) Eq. (45)

The other factors in OBJoverall include the objective function OBJ; a function of number of passes; and a correction on overall flood.The functionality of these other factors is given below.

1N1Nf p

+= (Customary or Metric) Eq. (46)

( )( ) 70% Flood Overall if

odOverallFlo70 odOverallFlof

70% Flood Overall if 1odOverallFlof<=

>=(Customary or Metric) Eq. (47)

The best overall design is defined as the one that minimizes OBJoverall.

DESIGN PRACTICES

NOMENCLATUREAb = Bubble area of pass, ft2 (m2) (see Figure 12 in Section III-A)Adi = Total downcomer inlet area for pass, ft2 (m2)Ado = Total downcomer outlet area for pass, ft2 (m2)Af = Free area for pass, ft2 (m2) (superficial area minus arithmetic average of inlet and outlet area of downcomer(s)

above the tray minus the waste area) (see Figure 13 in Section III-A)Ao = Hole area for pass, ft2 (m2)APC = Hole area correction term in jet flood equation, dimensionlessAs = Superficial (total) tower area, ft2 (m2)Aw = Waste area for pass, ft2 (m2) (normally zero for new sieve tray designs)b = Weir length term for Hofhuis tray clear liquid height calculation; see Eq. (31), in/ft2 (mm/m2)Cb = Capacity factor based on pass bubble area, VL/Ab, ft/s (m/s)CbF = Capacity factor at jet flood, ft/s (m/s)C4 = Intermediate to Eq. (24)cD = Droplet drag coefficient, dimensionlesscoe = Downcomer exit coefficient, see Text, Eq. (36)Cf = Capacity factor based on pass free area, ft/s (m/s)

Co = Pass vapor capacity factor based on hole area = 0.5

vo ρρ

−,ft/s (m/s)

Cs = Capacity factor based on tower superficial area, ft/s (m/s)Cult = Capacity factor at the tray ultimate capacity limit, ft/s (m/s)Cuniv = Capacity factor at the universal ultimate capacity limit, ft/s (m/s)c = Downcomer clearance between tray and downcomer apron at tray inlet, in. (mm) (see Figure 6B in

Section III-A)do = Hole diameter, in. (mm)dPdc = Two phase mixture pressure drop in the downcomer, in. (mm) of hot liquidDp = Drop size, in. (mm)Dopt = Optimum tower diameter as function of tray spacing & number of passes, ft (m)Dt = Tower diameter, ft (mm)EO = Overall efficiency, % (see Section lIl-l)fe = Fractional entrainment, dimensionlessff = Foaming factor, dimensionlessFF = Fouling factor, dimensionlessfw = Fractional weepage, dimensionless (Qw / QL)Fw = Weir constriction factor, dimensionlessgc = Gravitational constant, 32.2 ft/s2 (9.8 m/s2)H = Tray spacing, in. (mm)HB = Tray spacing below tray, in. (mm)hc = Hofhuis clear liquid height on tray, in. (mm) of hot liquidh'c = FRI TR 88 clear liquid height on tray, in. (mm) of hot liquidh''c = FRI TR 119 clear liquid height on tray, in. (mm) of hot liquidhcDP = Clear liquid height at dump point, in. (mm) of hot liquidh''cwp = FRI TR 119 clear liquid height at weep point, in. (mm) of hot liquidhd = Downcomer filling, in. (mm) of hot liquidhdc = Head loss due to two-phase flow through the downcomer, in. (mm) of hot liquidhed = Effective dry tray pressure drop, in. (mm) of hot liquidhf = Tray froth height, in. (mm) of hot liquid

hi = Tray inlet head, in. (mm) of hot liquidht = FRI TR 88 total tray pressure drop, in. (mm) of hot liquidh''t = FRI TR 119 total tray pressure drop, in. (mm) of hot liquidhud = Head loss under the downcomer or splash baffle corrected for froth flow, in. (mm) of hot liquidhudL = Liquid head loss under the downcomer or splash baffle, in. (mm) of hot liquidhvt = Head loss of vapor flowing through off-center vapor tunnel, in. (mm) of hot liquidhwi = Inlet weir height, in. (mm)hwo = Outlet weir height, in. (mm) (see Figure 3, Section III-A)KD = Correction factor for downcomer frictional pressure drop, dimensionlessKφ = Tray geometry factor in E-Method entrainment correlationKL = Liquid rate-tray spacing factor in E-Method entrainment correlationKσ = System properties factor in E-Method entrainment correlationKσµ = System properties factor in the low liquid rate E-Method entrainment correlationKε = Vapor energy dissipation term in E-Method entrainment correlationL = Liquid rate, gpm/in. of effective weir/pass, (dm3/s/m of weir/pass)L′ = Liquid rate, gpm/in. of diameter/pass (dm3/s/m of diameter/pass)LL = Liquid load, ft3/s (dm3/s) at conditionsLL(Min) = Minimum liquid load, ft3/s (dm3/s) at conditionsLw = Liquid rate weeping factorIc = Side downcomer chord length at inlet, in. (mm)Ic* = Inboard tray downcomer chord legnth at inlet, in, (mm)ldb = Pass length of downcomer bottom, in. (mm)lfp = Flow path length (distance between inlet and outlet downcomers), in (mm) (see Figure 12 in Section III-A)Ii = Pass inlet weir length, in. (mm)lo = Pass effective outlet weir length, in. (mm) (see Figure 12 in Section III-A)Io* = Pass outlet weir length on inboard tray, in. (mm) (see Figure 12 in Section III-A)lud = Pass length of bottom edge of downcomer or splash baffle, in. (mm) (see Figure 12 in Section III-A)NA = Number of actual traysNT = Number of theoretical traysNp = Number of passesn = Low liquid rate entrainment parameter [see Eq. (24) for mathematical definition]OBJ = Objective function for Dopt at given tray spacing and number of passes, dimensionlessOBJoverall= Overall objective function to determine best design, dimensionlessP = Calculated hole pitch, in. (mm)QDP = Pass weeping rate at the dump point, gpm (dm3/s) at conditionsQL = Pass liquid rate, gpm (dm3/s) at conditionsQw = Pass weeping liquid rate, gpm (dm3/s) at conditionsqv = Pass volumetric vapor rate, ft3/s (m3/s) at conditionsr = Downcomer inlet width, in. (mm). For modified arc downcomers, r represents the minimum pinch point rise, see

Section III-K rud = Downcomer bottom width, in. (mm)Rec = Critical Reynolds number, dimensionlessS5 = Dry tray pressure drop parameterS6 = Clear liquid height parameterS9 = Clear liquid height parametert = Tray thickness, in. (mm)TSmax = Tray spacing maximum used in optimization matrix, in. (mm)

DESIGN PRACTICES

TSmin = Tray spacing minimum used in optimization matrix, in. (mm)TSstep = Tray spacing step used in optimization matrix, in. (mm)uL = Liquid velocity across pass tray deck, ft/s (m/s)ut = Terminal velocity in universal ultimate capacity equations, ft/s (m/s)Vb = Vapor velocity through the pass bubble area, ft/s (m/s)Vbwp = Vapor velocity through the pass bubble area at the weep point, ft/s (m/s)Vdi = Velocity of clear liquid entering pass downcomer, ft/s (m/s)Vdi(ult) = Allowable downcomer entrance velocity limit, in. (mm)Vdo = Downcomer outlet velocity, ft/s (m/s)Vf = Vapor velocity based on average tower free area, ft/s (m/s)

VL = Pass design vapor load = 0.5

v ρρρq

− at conditions, ft3/s (m3/s)

VL(Ult) = Ultimate capacity vapor load dependent on system properties, ft3/s (m3/s) at conditionsVo = Pass vapor velocity through the holes, ft/s (m/s)Vud = Velocity under the downcomer, ft/s (m/s)Wec = Critical Weber number, dimensionlesswE = Pass liquid entrainment rate, lb/s (kg/s)wL = Pass liquid mass flow rate, k lb/hr (kg/s)wv = Pass vapor mass flow rate, k lb/hr (kg/s)w`v = Pass vapor mass flow rate at universal ultimate capacity limit, k lb/hr (kg/s)x = Multiplier on vapor and liquid rates in downcomer flood calculation, dimensionlessXL = Gpm/in of weir correction factor

α = Froth density on tray using FRI TR 119, dimensionless

β = Factor in ultimate capacity equation, 0.5

vLρρρ1.4

− (Customary and Metric)

∆ρ = Density difference, (ρL - ρv)Φ = Lockhart-Martinelli Flow Parameter, dimensionlessφ = Liquid head correction term in dry tray pressure drop, dimensionlessγ = Correction term for overall flood equationµL = Liquid viscosity at conditions, cP (mPa•s)µv = Vapor viscosity at conditions, cP (mPa•s)Ω = Weir height term in Hofhuis tray clear liquid height calculation; see Eq. (31), dimensionlessζ = Vapor fraction of aerated liquid flowing under downcomer (vapor carryunder), dimensionlessρL = Liquid density at conditions, Ib/ft3 (kg/m3)ρm = Vapor / Liquid mixture density under the downcomer = ( ) Lv ρζ1ρζ ⋅−+⋅ , Ib/ft3 (kg/m3)

ρv = Vapor density at conditions, Ib/ft3 (kg/m3)σL = Liquid surface tension at conditions, dynes/cm (mN/m)σSTD = Standard liquid surface tension, dynes/cm (mN/m) (see Figure 2)ψ = Downcomer froth density, fraction of froth volume occupied by liquid, dimensionless

COMPUTER PROGRAMS

AVAILABLE PROGRAMS

For up-to-date information on available programs and for user instructions, please contact your Technical ComputingRepresentative or consult with a FRACTIONATION SPECIALIST for help.

SOURCE PROGRAM NAME OR NUMBER VERSION NUMBER

* PEGASYS ExxonMobil Tower Internals Program 1.01PRO/II MoTIP is recommended until EMoTIP is

integrated with PRO/II5.55

The ExxonMobil Tower Internals Program utilizes the design equations contained in this section and the design limits in Table 1.EMoTIP is the recommended program to use for new tower or tray designs and for rating purposes. Existing tray designs can berated by specifying all of the tray hardware dimensions in "user specifies all geometry" mode.

* If Pegasys 5.2 is installed after September 1, 2003, the EMoTIP program will be available as part of the Pegasys install.

DESIGN PRACTICES

Table 1Sieve Tray Design Principles

(METRIC VALUES SHOWN IN PARENTHESES)

DESIGN FEATUREVALUES

SUGGESTEDALLOWABLE

RANGE COMMENTS1. Tray Spacing 15 to 36 in.

(350 to 900mm)

12 to 36 in.(300 to 900mm)

It is generally economical to use minimum values, as limited by downcomerfilling or maintenance considerations. Use of variable tray spacing toaccommodate loading changes from section to section should be consideredto minimize tower height and avoid tower swaging. Less than 15 inch (350mm) spacing is not recommended for new designs.

2. Number of LiquidPasses

1, 2 or 4 1, 2 and 4(3-passdesigns arenot available inEMoTIP)

EMoTIP design algorithm attempts to pick the correct number of passes. Italso develops alternate designs for every combination of tray spacing andnumber of passes where a feasible design can be achieved. These alternatedesigns are contained in the C:\TEMP\EMOTIP.DBG file after a design run.An old rule of thumb for selecting the number of passes is: for diameters 5 ft(1500 mm) and less, use single pass; for diameters over 5 ft (1500 mm), lookat 2 passes if the liquid rate exceeds about 7 gpm/in. of diameter (17 dm3/s/mof diameter) and try 1 pass if the liquid rate is equal to or less than 7 gpm/in.of diameter (17 dm3/s/m of diameter). For the final design, choose thenumber of passes which minimizes the total tower cost (i.e., tower height anddiameter). If the liquid rate exceeds 17.5 gpm/in of weir/pass (42.5 dm3/s/mof weir/pass) consult your FRACTIONATION SPECIALIST. The minimumdiameter for 4 pass trays is 10 ft (3050 mm).

3. Downcomers andWeirsa) Type of

downcomerChordal Segmental

[with 6 in. (150mm) minimumrise]

Inlet and outlet chord length must be at least 62.5% of the tower diameter forgood liquid distribution.

b) Center, off-centerdowncomer width(inlet and outlet)and anti-jumpbaffles

Inlet width: 8in. min.(200mm min.)Outlet width: 6in. min. (150mm min.)

Whenever the liquid rate exceeds 4.2 gpm/in. of diameter (10 dm3/s/m ofdiameter/pass), use a 14-16 in. (350-400 mm) high anti-jump baffle, sus-pended lengthwise in the middle of the center and off-center downcomer andextending the length of the downcomer. This will prevent froth from chokingthe downcomer as it converges from opposite sides. The base of the anti-jump baffle should be level with the top of the outlet weirs or the tray deck ifno weirs are present (see Figure 14 in Section III-A.)

c) Outlet weir height 2 in. (50 mm) 0 to 4 in. (0 to100 mm)

The optimum weir height is the one which maximizes tray efficiency withoutcreating downcomer sealing or filling problems. This optimum usually occursat a height of 2 to 3 in. (50 to 75 mm). See Section lII-l, Tray Efficiency, formore details.

d) Clearance underthe downcomer

1.5 in. (38 mm) 1 in. and up(25 mm andup)1.5 in. and up(38 mm andup) in foulingservices

Set the clearance to meet the downcomer sealing criteria and the velocityunder the downcomer criteria. Higher values of head loss under thedowncomer can be used if necessary to assure sealing of the downcomer, aslong as the maximum Vud criteria is not exceeded. If high liquid rates occur,consider use of a shaped downcomer to reduce the head loss. (See SectionIII-A, Figure 11.) However, do not use a shaped downcomer with a recessedbox or inlet weir.

e) Downcomer seal Operating orprocess seal(see SectionIII-A)

Inlet weir orrecessed inletbox should beavoided infoulingservices

In most cases, the liquid level on the inlet side of the tray can be made highenough to seal the downcomer through the use of the outlet weir (operatingseal). However, if the sum of the clear liquid height at the inlet to the tray (hi)and the head loss under the downcomer (hud) plus 0.5 in. (6 mm) is less thanthe downcomer clearance at design rates, the downcomer will not be sealed.Downcomer unsealing of up to 1.0 to 1.5 in (25 to 38 mm) at turndown in highliquid rate services are acceptable or in services where there is a need tominimize fouling. If the downcomer is unsealed, consider decreasing theclearance, increasing the outlet weir height or using an inlet weir or recessedinlet box. Inlet weirs add to downcomer filling; in some cases they may bedesirable for 3-pass or 4-pass trays to insure equal liquid distribution.Recessed inlet boxes are more expensive but may be necessary in caseswhere an operating seal would require an excessively high outlet weir.

Table 1 (Cont)Sieve Tray Design Principles

(Metric Values Shown in Parentheses)

DESIGN FEATUREVALUES

SUGGESTEDALLOWABLE RANGE

COMMENTS4. Hole Size and

Layouta) Hole diameter 1/2 in. (13 mm)

3/4 in. to 1 in. (19to 25 mm) forfouling services

1/8 to 1in. (3 to 25 mm) Smaller holes generate less entrainment for some situations,such as operation in the spray regime. However, holes smallerthan 3/8 in. (9 mm) in diameter on carbon steel trays may rustover during hydrostatic testing and should be avoided. Theuse of alloy trays to overcome this problem should beconsidered. Larger hole sizes should be used if fouling isanticipated.

b) Ratio of holearea to bubblearea (Ao/Ab),percent

6 to 10% 3.5 to 15% In general, the lower the open area, the higher the efficiencyand the lower the capacity. A tray with 8% open area givesgood efficiency and flexibility without a capacity debit for a widerange of liquid rates. Higher open areas may be required atvery low liquid rates to avoid the spray regime, at very highliquid rates to prevent downcomer filling, or in vacuum serviceto avoid excessive pressure drop. Percent hole areas below3.5% are not recommended because the distance betweenholes becomes excessive and liquid channeling may occur.(See discussion of blanking below.)

c) Bubble area, Ab 40 to 90% of As 40 to 90% of As Ab/As ratios below 40% or above 90% must not be used,because they are outside the range of available data. For trayshaving a significant amount of waste area, the Ab/As ratio isbased on dividing Ab by (As - Aw).

d) Hole blanking Blanking is generally not required unless the tower is beingsized for future service at much higher rates or if some trayshave much lower vapor loadings than the rest of the tower,(e.g., upper trays of absorber de-ethanizers and lower trays ofheavy hydrocarbon steam strippers). To maintain bestefficiency, blank uniformly within the bubble area. See GP 05-02-01 for more details on tray blanking.

e) Dry TrayPressure Drop

3.0 in. (76 mm) 1.25 - 5.5 in(32 - 140 mm)

Maximum allowable of 5.5 in. (140 mm); in foaming service2.25 in. (57 mm).

5. Tray Efficiency Calculate perSection III-I

Calculate per Section lIl-l Sieve tray efficiency will be equal to or better than that of othercommercially available trays provided there is not anentrainment or excessive weeping problem.

6. Foaming DesignCriteriaa) Set foam factor

to 1.3 or larger1.3 or greater Foam factor will derate both jet flood and downcomer flood. It

will also ensure a design with low downcomer inlet velocities ifthe downcomer choke is kept below 100%.

b) Maximum DryTray PressureDrop

2.25 in. (57 mm) Design for a dry tray pressure drop of 2.25 in. (57 mm) or less.If turndown becomes a problem, consider valve trays andconsult your FRACTIONATION SPECIALIST.

c) Radius tip anddowncomerclearance

1.5 in (38 mm)radius tip2.5 in (64 mm)downcomerclearance orlarger

Design using at least 1 in. (25 mm) radius tips. Also use agenerous downcomer clearance of 2.5 inches (64 mm) orlarger to allow foam to escape the downcomer. If turndowndowncomer seal becomes a problem consult yourFRACTIONATION SPECIALIST.

DESIGN PRACTICES

Table 2System Factors

PROCESS TYPE OF TOWER FOAMING FACTOR FOULING FACTOR

Powerformer Feed Stripper See Table A 1Deisopentanizer 1 1Depentanizer 1 1Stabilizer (deisobutanizer) Top 1 2

Bottom 1 1Absorber-Deethanizer 1.2 1Debutanizer Top 1 2

Bottom 1 1Splitter and Rerun Tower See Table A 1

Hydrotreater Kerosene Stripper See Table A 1Gas Oil Stripper See Table A 1

Polymers Propylene/Propane Splitter 1 1C2/Propylene Splitter - 1Hexane Undercutter 1 1Hexane Drying Tower Top - -

Bottom - 1Hexane Recovery Tower 1 1Isobutylene Purification Tower 1 1

FCCU FCCU Main Fractionator Top See Table A 2 Middle See Table A 1

Bottom See Table A 3Water Wash Section Top 1.2 2

Bottom - -Light Cat Naphtha Splitter Top 1 1

Bottom See Table A 1Stabilizer See Table A 1Debutanizer 1 1Absorber-Deethanizer 1.2 1Sponge Oil Absorber Top - -

Bottom See Table A 1Distillate Stripper Top - -

Bottom See Table A 1Crude Unit Prefractionator See Table A 2

Atmospheric Pipestill See Table A 2Vacuum Pipestill See Table A 2Sidestream Stripper (APS) See Table A 1Sidestream Stripper (VPS) See Table A 1Heavy Hydrocarbon Bottom Stripper See Table A 3Naphtha Splitter 1 1Depropanizer 1 1Debutanizer 1 1Deisopentanizer 1 1

Aromatics Deisohexanizer 1 1Benzene Tower 1 1Toluene Tower 1 1

PROCESS TYPE OF TOWER FOAMING FACTOR FOULING FACTORAromatics cont. Xylenes Splitter 1 1

C8/C9 Splitter 1 1Steam Cracker (See Note 2) Primary Fractionator See Table A 2

Demethanizer Prefractionator 1.3 0

C2 Topping Still (<320 psia) 1 0C2 Topping Still (>320 psia) 1.1 0Demethanizer Top 1.2 0 Bottom 1.15 0Ethylene/Ethane Splitter 1 0Deethanizer (btm. < 160F) Top 1.1 0 Bottom 1.1 1Deethanizer (btm > 160F & sig. C4 ) Top 1.1 1 Bottom 1.1 2Depropanizer Top 1 3 Bottom 1 3Debutanizer 1 3Absorber-Deethanizer Top 1.2 2

Bottom 1.2 3Absorber-Depropanizer Top 1.2 1

Bottom 1.2 3Steam Cracked Naphtha Rerun Tower Top 1 1

Bottom 1 2Steam Cracked Naphtha Stripper Top - -

Bottom 1 2Distillate Stripper (bottoms < 250F) 1 1Distillate Stripper (bottoms > 250F) Top 1 1

Bottom 1 2Debenzenizer 1 1Detoluenizer 1 1Primary Absorber Top - -

Bottom See Table B 1Sponge Oil Absorber Top - -

Bottom See Table B 1Caustic Treater 1.3 2

Solvents iC3OH Dehydration Tower - 1Hexane Tower 1 1Heptane Tower 1 1

Gofiner Product Stripper Top 1 2Bottom 1 1

Gas Treating H2S/MEA Absorber 1.3 2H2S/DEA Absorber 1.3 2H2S Regenerator - -Ucarsol Absorber 1.3 2

Water wash 1.1 1Ucarsol Regenerator 1.3 2

Water wash 1.1 1Sulfinol Absorber 1.3 2

Water wash 1.1 1

DESIGN PRACTICES

PROCESS TYPE OF TOWER FOAMING FACTOR FOULING FACTORGas Treating (Con't) Sulfinol Regenerator 1.3 2

Water wash 1.1 1Sour Water Stripper 1.3 2

Catacarb Absorber Top 1.4 2Bottom 1.3 1

Catacarb Regenerator 1.3 1Caustic Absorber 1.5 1Amine (FLEXSORB) Absorber 1.3 2Amine (FLEXSORB) Regenerator 1.3 2Glycol Regenerator 1.3 1Glycol Dehydrator 1.3 1Caustic Regenerator 2.5 1Selexol Absorber 1.3 1Selexol Regenerator 1.3 1

Miscellaneous Light Ends Dehexanizer 1 1Propylene/Propane 1 1C3/C4 Splitter 1 1i-Butane/n-Butane 1 1i-Butane/1-Butene 1 11-Butene/n-Butane 1 1Butadiene Purification Tower 1 1Cyclohexane/n-Heptane 1 1i-Octane/Toluene 1 1

Miscellaneous Towers Coker Main Fractionator Top 1 2 Middle 1 1

Bottom 1 4Hydrocracker Main Fractionator 1 2Quench Tower 1 1Isostripper (Alky Main Fractionator) 1 3Methanol/Water 1 1Ethanol/Water 1 1Isopropanol/Water 1 1Glycol Fractionator Top 1.2 1

Bottom 1.2 2ACN Extractive Distillation 1.4 1

LNG/NGL/LPG Condensate Stripper/Stabilizer 1 1Nitrogen Rejection 2 1Scrub Tower Top 1.1 1

Bottom 1 1Demethanizer Top 1.2 1

Bottom 1.1 1Deethanizer Top 1.1 1

Bottom 1 1Depropanizer 1 1Debutanizer 1 1Deisopentanizer 1 1

NGL Splitter 1 1

TABLE A

HYDROCARBON

TABLE B

ABSORBER

TABLE C

LOW SURFACE TENSION (SEE NOTE 1)

FoamingFactor

Molecular Weight(Note 3) Foaming Factor MW Lean Oil Foaming Factor

MultiplierSurface Tension(dyn/cm (mN/m)

1 <150 1.05 <100 1.05 2-2.51.05 150-250 1.1 100-150 1.1 1.5-21.1 250-300 1.2 150-300 1.2 1-1.51.2 >300 1.3 >300 1.3 <1

Notes 1: Apply "low surface tension foaming factor multiplier" only if a foam factor not already available from Table 2.2: The foam factors for ethylene recovery towers are being reevaluated. If applying these foam factors results in exceeding the allowable

tray design criteria please consult with a FRACTIONATION SPECIALIST.3: The information on the foam factor dependency on molecular weight is general guidance and may be conservative at times and it may be

overridden if a particular engineer has experience that shows otherwise.

Table 3Equations For Determining Liquid And Vapor Splits

The equations for liquid, vapor, and pressure drop distribution are presented here for three and four pass trays. The appropriateequations in each set are modified for the special case of vapor crossover. Each set of equations must be solved by trial-and-error. EMoTIP contains a convergence procedure that solves the flow rates and pressure drops for multipass trays. This programshould be used for all multipass tray designs. Note that feed trays, drawoff trays, and reboiler return areas of three and four passtrays can get complicated, and may not be thoroughly discussed in Section III-H. Therefore, consultation with aFRACTIONATION SPECIALIST is recommended.

Three Pass Trays

To determine the vapor and liquid flow rates for each pass of a three pass tray, the following equations must be solved (refer toFigure 8). This is a trial and error procedure for rating purposes only.

NO VAPOR CROSSOVERWITH VAPOR CROSSOVER

Replace Eqs. (4) and (5)

(1) QLa = QLC

(2) hdb = hdc

(3) QLa + QLb + QLc = QLtotal (4) If hta > htb, then hta = htb + hvt(4) wva = wvc If htb > hta, then htb = hta + hvt(5) hta + htc = 2 htb (5) htb = htc(6) wva + wvb + wvc = wvtotal

where:hta = heda + hca

htb = hedb + hcb

htc = hedc + hcc

Note: Sub-subscripts a, b, and c refer to the pass of the tray.

DESIGN PRACTICES

Four Pass Trays

To determine the vapor and liquid flow rates for each pass of a four pass tray, the following equations must be solved (refer toFigure 9). This is a trial and error procedure.

NO VAPOR CROSSOVERWITH VAPOR CROSSOVER

REPLACE EQS. (5), (6), (7), AND (8)

(1) QLa = QLc

(2) QLb = QLd

(3) hdc = hdd

(4) QLa + QLb + QLc + QLd = QLtotal (5) If hta > htb, then hta = htb + hvt(5) wva = wvc If htb > hta, then htb = hta + hvt(6) wvb = wvd (6) htc = htd(7) hta + htc = htb + htd (7) 2wva + 2wvb = wvtotal

(8) wva + wvb + wvc + wvd = wvtotal (8) 2wvc + 2wvd = wvtotal

where:hta = heda + hca

htb = hedb + hcb

htc = hedc + hcc

htd = hedd + hcd

Note: Sub-subscripts a,b,c, and d refer to the pass of the tray

Table 4Default Design Algorithm Values

Table 4A (Customary Units)

Default Values (1)

Min Max Step Target

Fouling and service-specific geometries

Hole diameter (do), in. Determined from fouling factor

Tray thickness (t), in. Determined from service

Geometry limits

Downcomer clearance (C), in. (2) 3.5 0.125

Downcomer top width (r), in. 6 H (3) 0.25

Downcomer bottom width (rud), in. (4) 6 r 0.25

1-pass: rud fracmin = rud / Diameter (5) 0.1097

Tray spacing (H), in. (2) 36 3

Flow path length (lfp), in. 16 180

Number of passes (Np) 1 4 (Allows 1,2,4)

Tower diameter (Dt), ft.: 1-pass 2.5 50 0.5

2-pass 5 50 0.5

4-pass 10 50 0.5

Hole area to bubble area ratio (Ao/Ab): sieve 0.035 0.15 0.0025

valve 0.05 0.18 0.0025

Effective weir length [for cases w/ picket fence], in. 0.25

Glitsch valve pitch, in. 3.0

Operating limits

Liquid flowrate per outlet weir length, gpm/in. 1.5 17.5

Universal ultimate capacity, % 80

Tray ultimate capacity, % OFmax(8)

Downcomer flood, % OFmax(8)

Jet flood, % OFmax+6(8)

Dry tray pressure drop, in. of hot liquid 1.25 5.5 (9)

Overall flood, % 80

Downcomer choke due to vapor bubble velocity, % 95

Froth to spray transition (% of PEGASYS limit) 110

Entrainment [both Exxon and Mobil models], % 10

Geometric mean of % DC Froth Backup and % DC Choke 70

Downcomer seal, in. -0.5

Weeping, % 20

Velocity under the downcomer, (Vud) ft/s 1.3 1.1

DESIGN PRACTICES

Notes: Table 4A(1) Allows user override of all geometry limits and operating limits (min, max, step, target).

(2) Determined from fouling factor and foaming factor.

(3) Large diameter towers are designed at the minimum rud / diameter given in table, and limited to a maximum of (r / H) = 1.20.

(4) Only straight downcomers are allowed when foaming factor ≥ 1.3.

(5) Based on 62.5% DC outlet length / diameter.

(6) Based on 60% side DC outlet length / nearest center DC chord length.

(7) Based on 60% side DC outlet length / nearest off-center DC chord length.

(8) OF designates overall flood.

(9) 2.25 if foaming factor ≥ 1.1.

Default Design Algorithm Values

Table 4B (Metric Units)

Default Values (1)

Min Max Step Target

Fouling and service-specific geometries

Hole diameter (do), mm Determined from fouling factor

Tray thickness (t), mm Determined from service

Geometry limits

Downcomer clearance (c), mm (2) 90 3.125

Downcomer top width (r), mm 150 H (3) 6.25

Downcomer bottom width (rud), mm. (4) 150 r 6.25

Tray spacing (H), mm (2) 900 75

Flow path length (lfp), mm 400 4500

Number of passes, (NP) 1 4 (Allows 1,2,4)

Tower diameter (Dt), mm: 1-pass 750 15,000 150

2-pass 1500 15,000 150

4-pass 3000 15,000 150

Hole area to bubble area ratio (Ao/Ab): sieve 0.035 0.15 0.0025

valve 0.05 0.18 0.0025

Effective weir length [for cases w/ picket fence], mm 6.25

Glitsch valve pitch, mm 75

Operating limits

Liquid flowrate per outlet weir length, dm3/s/m 3.73 43.47

Universal ultimate capacity, % 80

Tray ultimate capacity, % OFmax(8)

Downcomer flood, % OFmax

Jet flood, % OFmax+6

Dry tray pressure drop, mm of hot liquid 31.75 139.7 (9)

Overall flood, % 80

Downcomer choke due to vapor bubble velocity, % 95

Froth to spray transition (% of PEGASYS limit) 110

Entrainment [both Exxon and Mobil models], % 10

Geometric mean of % DC Froth Backup and % DC Choke 70

Downcomer seal, mm -12.7

Weeping, % 20

Velocity under the downcomer, m/s 0.396 0.335

DESIGN PRACTICES

Notes: Table 4B(1) Allows user override of all geometry limits and operating limits (min, max, step, target).

(2) Determined from fouling factor and foaming factor.

(3) Large diameter towers are designed at the minimum rud / diameter given in table, and limited to a maximum of (rud / H) = 1.20.

(4) Only straight downcomers are allowed when foaming factor ≥ 1.3.

(5) Based on 62.5% DC outlet length / diameter.

(6) Based on 60% side DC outlet length / nearest center DC chord length.

(7) Based on 60% side DC outlet length / nearest off-center DC chord length.

(8) OF designates overall flood.

(9) 57.15 if foaming factor ≥ 1.1.

Figure 1Weeping And Dumping Regions

Weeping Region

Vapor Rate

DumpPoint

DumpingRegion

UnacceptableEfficiency

FairTo Poor

EfficiencyGood

Efficiency

DP3BF10

WeepPoint

DESIGN PRACTICES

Figure 2EMoTIP Tray Performance Diagrams

(Customary Units)

Figure 2A EMoTip Performance Diagram For C6/C7, 24 PSIA

Physical properties: ρV = 0.321298 lb/ft3

ρL = 40.7 lb/ft3

µV = 0.0082 cP µL = 0.239 cP σ = 13.91 dyn/cm ff = 1.00

0 2 4 6 8 10 12 14 16

Liquid rate (gpm / inch of weir)

C b (f

Figure 2B EMoTIP Performance Diagram For iC4/nC4, 165 Psia

ρL = 30.7 lb/ft3

0 2 4 6 8 10 12 14 16

C b (f

DESIGN PRACTICES

Figure 2C EMoTIP Performance Diagram For iC4/nC4, 300 PSIA

ρL = 27.1 lb/ft3

0 2 4 6 8 10 12 14 16

C b (f

Figure 3E-Method Entrainment Kφφφφ Factor

(Same for Customary and Metric Units)

0.00 0.05 0.10 0.15 0.200.01

.080.1

0.60.81.0

6.08.010

400500

Ao / Ab DP3BF04

Kφ = (1*109)(J5)3.28 (Ao / Ab)6.12 / (do / t)0.76

do / t = 6-16

Note: Obtain J5 from

J5 = KD 0.4 (1.25 (Ao / Ab)) + (1-(Ao / Ab))2

Where : KD = 0.1492 + 0.0776 In (do / t)

DESIGN PRACTICES

Figure 4E-Method Entrainment Kl Factor

Figure 4A (Customary Units)

(H, in.)

KL = 1.22*106 (L1.08 / H5.23)

0.015 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

L, gpm / in. of Weir / Pass

DP3BF5A

Figure 4BE-Method Entrainment Kl Factor

(Metric Units)

300Tray Spacing, mm

KL = 0.00207 (L1.08 / (H / 1000)5.23)

6050453530252015105

L, dm3 / s Per Meter of Weir / Pass

.080.1

0.40.50.6

405060

DP3BF5B

DESIGN PRACTICES

Figure 5E-Method Entrainment Kσσσσ Factor

(Same for Customary and Metric Units)1,000

600500

0.51 2 5 10 20 50 100

σ L, dynes / cm (mN / m)

Kσ = 1982 σ L1.85

DP3BF06

Figure 6E-Method Entrainment Kεεεε Factor

Figure 6A (Customary Units)

1x103 1x102 1x1011x1041x1051x106

Kε = 1 x 10-12 (Vo2 wv / Ab)

400500600

8001000

30004000500060008000

400005000060000

80000100000

1x101 1 1x101 1x102 1x103 1x104

wv / A

DP3BF7A

DESIGN PRACTICES

Figure 6BE-Method Entrainment Kεεεε Factor

(Metric Units)

1x106 1x105 1x1041x1071x1081x109

405060

300400500600

8001000

400050006000

800010000

1x104 1x103 1x102 1x101 1 10

V o2 w

DP3BF7B

Kε = 1.21 x 1012 (Vo2 wv / Ab)3.28

Figure 7Kσσσσµµµµ Factor For E-Method Entrainment Correlation

(Same for Customary and Metric Units)

Surface Tension (σ), dynes/cm (mN/m)DP3Bf02

1 10 100

µL = 0.05

Legend:Liquid Viscosity (µL), cP (mPa s)

σSTD = 10

σLσSTD

Kσµ = 1.0 for ≥ 1.0σL

σSTDfor < 1.0Kσµ =

σLσSTD

1.68 - (0.244/µL0.55)

DESIGN PRACTICES

Figure 8Three-Pass Tray Geometry

Pass CPass A Pass B

h wo a

h wo b

h wo c

Pass C Pass B Pass Ah w

h wi b h w

cb ccca

7 6 345 2 1

Note: The numbers shown on the plan view refer to the dimensions required as input to the 1143 computer program. Sub-subscripts a,b, and c refer to the pass of the tray.

Figure 9Four-Pass Tray Geometry

1 2 3 4 5 5 4 3 2 1

10 9 8 7

6109876

cC cd cCcd

h wi c

ca cacbcb

h wi a

h wi b

h wi a

Note: The numbers shown on the plan view refer to the dimensions required as input to the 1143 computer program. Sub-subsripts a, b, and c refer to the pass of the tray

DESIGN PRACTICES

ExxonMobil Research and E

Figure 10EMoTIPSieve And Valve Tray Design Algorithm

Initialization

Tray spacing loop(Start with Hmin)

Number of passes loop(Start with Npmax)

Diameter loop(Start with Dtmin)

Downcomer top widthloop (Start with rmin)

Downcomer bottomwidth loop (Start with

rud min)

Ao/Ab loop (Start withAo/Abmin)

Rate tray; apply OBJ

Ao/Ab = Ao/Abmax?

rud =rudmax?

Ao/Ab=Ao/Ab+Ao/Abst

r =rmax?

rud=rud+rstep

Any good design at Dt?

r =r+rstep

NP =NPmin?

Dt =Dt+Dtstep

H =Hmax?

Reduce Np (4 to 2 to 1)

H =H+Hstep

Optimum designfound

Calculate OBJoverall formatrix of designs

ngineering Company – Fairfax, VA

Figure 11Dry Tray Pressure Drop Design Consideration Function

(CUSTOMARY UNITS)

Dry Tray Pressure Drop Factor for Objective Function

1.25 1.75 2.25 2.75 3.25 3.75 4.25 4.75 5.25

hed, in of hot liquid

f(hed)=1+(hed-3.5)2/100

DESIGN PRACTICES

Figure 12Liquid Load Design Consideration Function

(CUSTOMARY UNITS)

Outlet Weir Loading Factor for Objective Function

1.5 3.5 5.5 7.5 9.5 11.5 13.5 15.5 17.5L, Gpm/in, of Weir/Pass

f(Gpm/in) = (1.9e-8)(Gpm/in)6 - (2.13e-6)(Gpm/in)5 + (8.59e-5)(Gpm/in)4 -(1.66e-3)(Gpm/in)3 + 0.017(Gpm/in)2 - 0.08(Gpm/in) + 1.14

Figure 13Froth/Spray Transition Design Consideration Function

Froth/Spray Transition Factor for Objective Function

0 20 40 60 80 100

Froth/Spray Transition, % of Exxon Max Allowed

f (FS) = 1 if FS < 70% = 1 + exp [ (FS-70) / 15 ] - 1 / 350 if FS > 70%

DESIGN PRACTICES

Figure 14Downcomer Choke Design Consideration Function

Downcomer Choke Factor for Objective Function

0 10 20 30 40 50 60 70 80 90 100Choke, %

f (Choke) = 1 if Choke < 40% = 1 + exp [ (Choke-40) / 15 ] - 1 / 550 if Choke > 40%

Figure 15Entrainment Design Consideration Function

Entrainment Factor for Objective Function

0 2 4 6 8 10 12 14 16 18 20E-Method + M-Method, %

f (Ent) = 1 + exp [ (E-Method + M-Method) / 5 - 1 ] / 1100

Figure 16Flow Path Length Design Consideration Function

Flow Path Length Factor for Objective Function

0 20 40 60 80 100 120 140 160 180 200lfp, in

f(lfp) = -0.0071(lfp)+1.21 if lfp < 30% = 1 if 30% <lfp < 70% = 0.000909(lfp)+ 0.9364 if lfp > 70%

DESIGN PRACTICES

Figure 17Weeping Less Than 20% @ Turndown

Turndown Cond. #1 Weeping Factor for Objective Function (Sieve and Fixed Valve)

0 2 4 6 8 10 12 14 16 18 20Weep, %

F(Weep) =1 + [exp (Weep/5-1)] /1100

ExxonMobil Research and Engineering Com

Figure 18Sealing Factor @ Design Rates Design Consideration Function

Design Cond. Sealing Factor for Objective Function

-1.25 -1.15 -1.05 -0.95 -0.85 -0.75 -0.65 -0.55Downcomer Seal, in.

f(Seal) = 1 + 0.5 * ( DCSealminallow - [DCSeal at design cond] ) only applied if DCSeal < DCSealminallow

Figure 19Weeping Rate Design Consideration Function

Turndown Cond. #2 Weeping Factor for Objective Function (Sieve and Fixed Valve)

0 5 10 15 20Weep, %

p)DESIGN PRACTICES

f(Weep2) =1 if Weep <20 at T/D =2 if Weep >20 at T/D

pany – Fairfax, VA

25 30 35 40

Figure 20Sealing Factor @ Turndown Rates Design Consideration Function

Turndown Cond. Sealing Factor for Objective Function

-1.25 -1.15 -1.05 -0.95 -0.85 -0.75 -0.65 -0.55 -0.45 -0.35 -0.25Downcomer Seal, in.

f(Seal2) = 2 + ( DCSealminallow - [DCSeal at design cond] ) only applied if DCSeal < DCSealminallow

Figure 21Vapor Carryunder Design Consideration Function

Vapor Carryunder Factor for Objective Function

0 2 4 6 8 10 12 14 16 18 20Percent Vapor Carryunder

SAMPLE PROBLEMA grassroots depentanizer is to be designed.After running a PRO/II simulation for a depentanizer, it is determined that 16 theoretical stages are required to obtain the desiredseparation, with 7 stages above the feed (rectifying section) and 9 stages below the feed (stripping section). In addition, thestripping section (below the feed) limits the tower capacity. The highest loaded tray in the stripping section, theoretical stage 16,has the following design flowrates and conditions according to the PRO/II model:

Operating pressure = 66.4 psiaLiquid temperature = 366.8 FLiquid rate = 560.82 klb/hr Vapor rate = 322.58 klb/hrLiquid density = 217.2 lb/bbl Vapor density = 0.972 lb/ft3Liquid viscosity = 0.1765 cP Vapor viscosity = 0.0096 cPSurface tension = 8.94 dyn/cmLiquid molecular weight = 125.9 Vapor molecular weight = 123.3

A turndown of 50% is required. This means that all trays must be designed so that they can handle both the design loads and50% of the design loads. For theoretical stage 16, a 50% turndown will be specified for design, but after designing thestripping section trays based on stage 16, it is necessary to check the minimum loaded tray in the section to ensure that it toocan handle 50% of design rates without significant loss of efficiency.

1) It is assumed that sieve trays have already been determined to be the correct internals selection for this tower. The loadings,conditions, physical properties, and turndown requirements for theoretical stage 16 are entered into EMoTIP. In addition, theequilibrium slope from the McCabe-Thiele diagram has been determined from the composition profile in the PRO/II outputreport to be 0.700; this value is entered into EMoTIP for calculating efficiency. The following output report is obtained afterrunning EMoTIP in design mode, using the default design algorithm limits:

DESIGN PRACTICES

**************************************************************************

***** EXXONMOBIL USE ONLY Page No. 1

**************************************************************************

***** EXXONMOBIL TOWER INTERNALS PROGRAM *****

***** VERSION 1.0 01-May-2003 *****

**************************************************************************

Date: 02-May-03 Time: 10:48:16

User Name: sample user Tower Tag No: sample tag

Project: Depentanizer sample Facility: sample facility

Comments: Depentanizer stripping section design based on theo 16

Process: Powerformer

Service: Depentanizer

Tower Section: Bottom

Sieve Tray 2 Pass Design Case

Tower Diameter = 11.00 ft Tray Spacing = 18.0 in

--------------------------------------------------------------------------

TRAY PERFORMANCE SUMMARY (EACH VALUE IS FROM ITS LIMITING PASS)

--------------------------------------------------------------------------

PRIMARY DESIGN PARAMETERS CRITERIA

--------------------------------------------------------------------------

Overall Percent Flood 77.4 % < 85.0 %

Percent Ultimate Capacity Tray 45.7 %

Percent Jet Flood 78.7 %

Percent Downcomer Flood 77.4 %

Probability of Non-Flooding Design 99.2 %

Turndown: % Thruput at 20 % Weep 40.4 %

Design Overall Effic. Key Comp 1 88.0 %

--------------------------------------------------------------------------

SECONDARY DESIGN PARAMETERS CRITERIA

--------------------------------------------------------------------------

Liquid Entrainment (M-Method) 0.01 % < 10.0 %

Liquid Entrainment (E-Method) 0.14 % < 10.0 %

DC Choking Due To Velocity 50.2 % < 100.0 %

SQRT(DC Backup * DC Choke) 58.7 % < 70.0 %

Liquid Rate, US gpm/in of weir 9.3 1.5 - 17.5

Froth/Spray Transition 55.0 % < 110.0 %

Ultimate Capacity Universal 38.5 % < 85.0 %

Dry Tray Pressure Drop, in 1.9 1.25 - 5.50

Total Tray Pressure Drop, in 5.3

Velocity Under Downcomer, ft/s 1.04 < 1.30

Vapor Fraction Under DC 0.04

Downcomer Seal in 0.39 > -0.50

Liquid Weeping at Conditions 0.0 % < 20.0 %

--------------------------------------------------------------------------

WARNING: See Page 6 for 5 warning messages.

DESIGN PRACTICES

--------------------------------------------------------------------------

FOR CONTRACTOR AND VENDOR USE Page No. 2

--------------------------------------------------------------------------

TRAY GEOMETRY SUMMARY * INDICATES INPUTTED GEOMETRY

--------------------------------------------------------------------------

* Tray Type = Sieve Avg. Adi = 10.054 ft2 10.6 %

No. of Pass = 2 Avg. Ado = 5.385 ft2 5.7 %

Diameter = 11.00 ft Tot. BA = 62.692 ft2 66.0 %

Tray Spacing = 18.0 in Tower Area = 95.033 ft2

PASS A PASS B

FLOW DIRECTION CNTR-SIDE SIDE-CNTR

* DC Type Sloped Sloped

DC Top Area ft2 10.071 10.036

DC Top Width in 21.500 11.000

DC Btm Area ft2 5.392 5.378

DC Btm Width in 14.000 5.875

Wasted DC Area ft2 0.000 0.000

Bubble Area ft2 32.067 30.625

Bubble Waste Area ft2 0.000 1.464

Free Area ft2 39.809 39.785

Volum. Waste Area ft2 0.000 0.000

Flow Path Length in 38.63 38.50

DC Top Chord Len in 97.48 130.15

DC Btm Chord Len in 81.29 131.48

DC Clearance in 2.500 2.125

Outlet Weir Ht in 2.375 3.500

Tray Thickness in 0.074 0.074

Hole Diameter in 0.5000 0.5000

Number of Holes 2057 1965

Hole Area ft2 2.806 2.680

Hole/Bubble Area % 8.75 8.75

NON-STANDARD TOWER INTERNALS

DC Lip Radius in 0.00 0.00

Inlet Weir Ht in 2.50 0.00

Inlet Weir Distance in 2.50 0.00

Recessed Inlet Box No No

Picket Fence Weir No Yes

Swept Back Weir No

Eff Weir Length % 100.00 74.90

Eff Weir Length in 97.48 97.48

Eff DC Btm Length % 100.00 100.00

Eff DC Btm Length in 81.29 131.48

Eff Flow Path Len % 100.00 100.00

Anti-Jump Baffle Req'd

NOTE: Values for DC areas, widths, and waste area for Pass B

are 50% of the total for the associated center DC.

NOTE: The specified pass for an inlet weir is the

pass upstream of the downcomer it seals.

--------------------------------------------------------------------------

LOADING AND PHYSICAL PROPERTIES

--------------------------------------------------------------------------

Service: Depentanizer

Pressure = 66.40 psia

Temperature = 366.80 degF

Liquid Flow = 1807.43 US gpm Vapor Flow = 92.22 ft3/s

= 560.8 klb/h = 322.6 klb/h

Liquid T/D Fact = 50.00 % Vapor T/D Fact = 50.00 %

Liquid Density = 217.200 lb/bbl Vapor Density = 0.972 lb/ft3

Liquid Viscosity = 0.176 cP Vapor Viscosity = 0.010 cP

Surface Tension = 8.941 dyn/cm

Liquid Mol Wt = 125.903

Foaming Factor = 1.00 Fouling Factor = 1 (Clean)

Defaults for this tower and section: foam fac = 1.00 foul fac = 1

DESIGN PRACTICES

--------------------------------------------------------------------------

EXXONMOBIL USE ONLY Page No. 3

--------------------------------------------------------------------------

TRAY HYDRAULICS

--------------------------------------------------------------------------

PASS A PASS B

PRIMARY PERFORMANCE INDICATORS

Overall Flood % 77.45 76.56

Ult Capacity Tray % 45.69 45.72

Jet Flood % 76.84 78.66

DC Flood % 77.38 66.35

Prob. of Non-Flood % 99.16 99.35

T/D to 20% Weep % 40.04 40.39

SECONDARY PERFORMANCE INDICATORS

Liquid Ent M-Method % 0.01 0.01

Liquid Ent E-Method % 0.09 0.14

DC Choke % 50.01 50.18

SQRT(DCBackup*Choke)% 58.75 49.39

Froth/Spray Trans % 52.48 54.96

Ult Capacity Univ % 38.52 38.52

Dry Tray Press Drop in 1.91 2.00

Tot Tray Press Drop psi 0.106 0.119

Velocity Under DC ft/s 1.43 1.04

Downcomer Seal in 6.27 0.39

Vapor Frac Under DC 0.04 0.04

Liquid Weep % 0.00 0.00

TURNDOWN PERFORMANCE INDICATORS

Downcomer Seal in 4.29 -0.11

Liquid Weep % 4.01 3.44

MISCELLANEOUS CALCULATIONS

--------------------------

JET FLOOD PARAMETERS

Liquid Load US gpm/in 9.270 9.270

C-Fact Based on BA ft/s 0.231 0.242

DOWNCOMER BACK-UP (IN OF HOT LIQUID)

FRI Tray Cl Liq Ht in 2.83 3.34

Tot Tray Press Drop in 4.74 5.34

DC Fric Head Loss in 0.73 0.85

Head Loss Under DC in 1.49 0.66

Tray Inlet Head in 4.78 1.86

Hofhius Tray CLH in 1.86 1.91

Tray Froth Density 0.271 0.257

Tray Froth Height in 6.85 7.44

DC Clear Liq Height in 11.92 8.87

DC Froth Density 0.848 0.849

DC Froth Height in 14.06 10.45

DC Backup % 69.01 48.61

Note: Downcomer calculations use Hofhius clear liquid height equation.

Total tray pressure drop uses FRI clear liquid height equation.

Percent Thruput at

85% Overall Flood (Const L/V) 108.703 109.070

Liq Vel into DC Top ft/s 0.200 0.201

DESIGN PRACTICES

--------------------------------------------------------------------------

TRAY EFFICIENCY CALCULATIONS

--------------------------------------------------------------------------

KEY COMPONENT NUMBER 1

DESIGN OVERALL EFFICIENCY: 87.97 %

EFFICIENCY MODEL USED: M-Method

TYPICAL DESIGN OVERALL EFF: 75 % FOR: Depentanizer

PHYSICAL PROPERTIES USED ONLY IN EFFICIENCY CALCULATIONS

Slope of Eqm Line = 0.700

Liquid Molecular Weight = 125.90

Vapor Molecular Weight = 123.28

Vapor Diffusivity = 0.8244E-05 ft2/s (Calculated)

Liquid Diffusivity = 0.8627E-07 ft2/s (Calculated)

PASS A PASS B

MASS TRANSFER PARAMETERS

Lambda 0.41 0.41

NOG 1.75 1.82

Liquid Phase Control

(from E-Method) % 14.07 13.52

Clear Liquid Height in 2.31 2.69

Froth Density 0.22 0.22

EFFICIENCY CALCULATIONS

Vert Mixing Pools, j 100.00 100.00

Point Efficiency % 82.31 83.54

Horiz Mixing Pools, N 7.88 5.77

Uncorrected Murphree

Tray Efficiency, EMV % 101.91 102.39

Lng FPL/Sml DBA Corr. 1.00 1.00

Weepage Correction 1.00 1.00

Entrainment Corr. 1.00 1.00

Corrected Tray Effic. % 101.90 102.37

Pass Averaged Tray Eff. % 102.14

Effective Tray Effic. % 102.14

Overall Lambda 0.41

Expected Overall Effic. % 103.50

Design Overall Effic. % 87.97

Note: 0.85 design safety factor included in Design Overall Efficiency

E-Method DESIGN OVERALL EFFICIENCY: 76.61 %

Note: The E-Method efficiency model is not suggested for this service.

DESIGN PRACTICES

--------------------------------------------------------------------------

COMPLETE LIST OF WARNING MESSAGES

--------------------------------------------------------------------------

Please see your local Fractionation Specialist regarding assistance with

your tray design and advice regarding warning messages and intrepreting

the output. Technology ownership of this program is the Fractionation

Technology Group in the CDFS section located at the Central Engineering

Office in Fairfax, VA.

WARNING: PAGE 4 OF THE OUTPUT REPORT (HYDRAULICS RATIOS) ONLY RELEVANT

FOR 3-PASS AND 4-PASS TRAYS. PAGE 4 NOT PRINTED FOR THIS CASE.

WARNING: OUTLET WEIR HEIGHT IS GREATER THAN 1/6 OF THE TRAY SPACING.

WARNING: ANTI-JUMP BAFFLES ARE REQUIRED ON CENTER DOWNCOMERS WHEN

LIQUID LOAD > 4.20 US gpm/in of weir .

WARNING: USER SHOULD FINE-TUNE DESIGNS TO MINIMIZE THE NUMBER OF

WARNING MESSAGES BY MAKING GEOMETRY ADJUSTMENTS IN RATING

MODE FOR NON-RATING OPTIONS.

WARNING: PICKET FENCE WEIRS APPLIED DUE TO LOW LIQUID RATE, OR TO

BALANCE THE PASSES OF 2-PASS OR 4-PASS TRAYS. CONTACT A

FRACTIONATION SPECIALIST BEFORE APPLICATION OF PICKET FENCE

WEIRS.

The design algorithm has determined that a 11.0 ft diameter, 2-pass tray with 18 inch tray spacing is the best design for thestripping section. This determination is based on a pseudo-cost objective function in the EMoTIP design algorithm. However, thebalance between diameter, number of passes, and tray spacing cannot be accurately captured in a simplified cost objectivefunction. Many factors, such as design of other sections of the tower, maximum total tower height, area available for the towerfootprint, and maintenance considerations can play a very important role in determining the optimum tower design.

For this reason, a design debug (*.DBG) file is created when a design case is run, that allows the user to compare alternatedesigns. The present design case gives the following design comparison table at the end of the debug file:

Optimum Diameters, ft

Num Pass 4 2 1

Tray Spacing, in 36.0 10.0 9.5 10.0

Tray Spacing, in 33.0 10.0 9.5 10.0

Tray Spacing, in 30.0 10.0 9.5 10.5

Tray Spacing, in 27.0 10.0 10.0 10.5

Tray Spacing, in 24.0 10.0 10.0 11.0

Tray Spacing, in 21.0 10.5 10.5 13.5

Tray Spacing, in 18.0 10.5 11.0 *****

For new designs, 18 inch tray spacing may be insufficient to allow for easy maintenance. If the designer wanted to increase trayspacing from 18 inches, the table above shows the minimum tower diameter needed to meet all of the design limits as a functionof tray spacing. Perhaps a 10.0 ft diameter, 24-inch tray spacing tower is the best design for this application. Or, the designercould decide to use 1-pass trays, in which case an 11.0 ft diameter, 24 inch tray spacing may be the best choice.

Another option available to the designer is the ability to specify design limits. Default limits are built into the program, and arerecommended for typical designs for which no additional information is known.Geometry design limits may be set if certain geometric constraints are known (for example, if the tower cannot be greater than 12ft in diameter).Performance design limits may also be set by the user, if the user wishes to make the tower more or less conservative than thedefault. This option is recommended for advanced users only.A design case with the default design limits can take several minutes to run, because the algorithm searches the entire designspace for the best design (the sample depentanizer design looked at over 2.6 million tray geometries). Limiting the design windowcan save time by reducing the number of tray geometries considered in the design algorithm.

2) Assuming that the 2-pass, 11.0 ft diameter, 18 inch tray spacing design is chosen, the design efficiency for the strippingsection design is 87.97% as given on page 5 of the output report. (9 theoretical stages) / (0.8797) = 10.2 actual trays. This isrounded up to 11 actual trays for the depentanizer stripping section.

3) For design of the depentanizer rectifying section, a similar process is performed as above. However, it must be determinedwhether a swedged tower is appropriate. This is not a large tower, so economics would likely justify keeping the samediameter for the stripping and rectifying section.

4) As mentioned in step 1, after designing each section, the least loaded tray must be rated at turndown conditions to ensurethat the tray will not exhibit excessive weeping, downcomer unseal, or very low dry tray pressure drop at turndown rates.

5) For comparison, the design results of sieve tray program 1133 in Pegasys 5.2 (the older tower internals program) aredisplayed below for the stripping section.

DESIGN PRACTICES

WARNINGS AND ERROR MESSAGES

---------------------------

HOLE DIAMETER NOT SPECIFIED, DEFAULT TO 0.5 IN (12.7 MM)

DESIGN LIQUID RATE DENSITY ASSUMED TO BE #/FT3

***** CAUTION ***** DESIGN CASES UNDER THE FOLLOWING CONDITIONS

MAY NOT BE OPTIMUM

AN ANTI-JUMP BAFFLE MUST BE PROVIDED ON THE INBOARD

DOWNCOMER IF THE LIQUID RATE EXCEEDS 4.2 GPM/IN

(10 DM3/S/METER) OF DIAMETER/PASS

SIEVE TRAY DESIGN PROGRAM NUMBER 1133

VERS. 7.5

TOWER DIAMETER FT 11.00

TRAY SPACING INCHES 18.000

NO. OF LIQUID PASSES 2.

HOLE AREA PER TRAY FT2 4.8325

FRACTIONAL WEEPAGE (MAX= 0.20) 0.032

DC FILLING, % (MAX= 50.0) 50./ 47.

* JET FLOOD, % (MAX= 90.0) 85./ 76.

ULTIMATE CAPACITY, % (MAX= 90.0) 44./ 46.

SPRAY TRANSITION, % (MAX=100.0) 49./ 53.

ENTRAINMENT, % (MAX= 20.0) 0./ 0.

++ DC INLET VEL FT/SEC (MAX= 0.516) 0.307 / 0.502

DC OUTLET VEL FT/SEC (MAX= 0.600) 0.307 / 0.502

DC INLET CHOKING (MAX= 1.0) 0.287 / 0.765

DOWNCOMER SEAL INCHES (MIN=-0.25) 0.310 / 0.572

FINAL TRAY DESIGN

-----------------

TOWER DIAMETER FT 11.00

TRAY SPACING INCHES 18.00

NO. OF LIQUID PASSES 2.00

HOLE SIZE INCHES 0.500

HOLE AREA PER TRAY FT2 4.8325

NO. OF HOLES 3544.

TRAY DECK THICKNESS INCHES 0.074

OUTBOARD INBOARD

DC INLET RISE INCHES 16.000 8.750

DC INLET AREA FT2 6.555 8.015

CHORD LGTH AT TOP OF DC INCHES 86.171 131.710

DC OUTLET RISE INCHES 16.000 8.750

DC OUTLET AREA FT2 6.555 8.015

CHORD LGTH AT BTM OF DC INCHES 86.171 131.710

DC CLEARANCE INCHES 1.500 1.500

RECESSED BOX NO NO

SHAPED DC LIP YES YES

DOWNCOMER TYPE CHORDL

OUTLET WEIR HEIGHT INCHES 1.500 1.500

DESIGN PRACTICES

INLET WEIR HEIGHT INCHES 0.000 0.000

CROSS SECTIONAL AREA FT2 95.033

FREE AREA FT2 87.018 81.923

WASTE AREA FT2 0.000 0.000

BUBBLE AREA FT2 73.908 73.908

HOLE/BUBBLE AREA PCT 6.5 6.5

BUBBLE/CROSS SECT AREA PCT 77.8 77.8

FLOW PATH LENGTH FT 3.802 3.802

VAPOR - LIQUID RATES AND PROPERTIES AT CONDITIONS

-------------------------------------------------

KILOLBS/HR OF VAPOR (DESIGN/MIN) 322.580/ 161.290

LB/FT3 OF VAPOR AT COND (DES/MIN) 0.9716/ 0.9716

VAPOR VISCOSITY AT COND CP 0.0096

FT3/SEC OF VAPOR AT COND 92.2247

VAPOR LOAD AT COND FT3/SEC 14.8028

TRAY LIQUID TEMPERATURE DEGF 366.8000

OPERATING PRESSURE PSIA 66.4000

KILOLBS/HR OF LIQUID (DESIGN/MIN) 560.8200 / 280.4100

LB/FT3 OF LIQUID AT COND (DES/MIN) 38.6850 / 38.6850

LIQUID RATE (DESIGN/MIN) US GAL/MIN 1807.3064 / 903.6532

SURFACE TENSION AT COND DYNES/CM 8.941

LIQUID VISCOSITY AT COND CP 0.176

SYSTEM TYPE NON-FOAMING HYDROCARBON

$ DOWNCOMER FILLING CALCULATIONS (INCHES ARE OF LIQUID AT CONDITIONS)

------------------------------

OUTBOARD/INBOARD

DRY TRAY PRESSURE DROP (HED) INCHES 3.30/ 3.30

CLEAR LIQUID HEIGHT (HC) INCHES 1.88/ 1.66 ##

TOTAL TRAY PRESSURE DROP (HT) INCHES 5.18/ 4.96

TOTAL TRAY PRESSURE DROP (HT) PSI 0.12/ 0.11

INLET HEAD (HI) INCHES 1.66/ 1.88

DC HEAD LOSS (HUD) INCHES 0.98/ 0.42

DC FILLING(DENSITY CORR) (HD) INCHES 8.98/ 8.40

DC FILLING, % (50.00 MAXIMUM) 49.90/ 46.68

DOWNCOMER VELOCITY CALCULATIONS

-------------------------------

++ INLET VELOCITY, FT/SEC (0.516 MAXIMUM) 0.307/ 0.502

OUTLET VELOCITY, FT/SEC (0.600 MAXIMUM) 0.307/ 0.502

DC INLET CHOKING (1.00 MAXIMUM) 0.287/ 0.765

TRAY CAPACITY CALCULATIONS

--------------------------

* JET FLOOD, % ( 90.0 MAXIMUM) 85./ 76.

ULTIMATE CAPACITY, % ( 90.0 MAXIMUM) 44./ 46.

SPRAY TRANSITION, % (100./100. MAX) 49./ 53.

ENTRAINMENT, % ( 20.0 MAXIMUM) 0./ 0.

TRAY FLEXIBILITY CALCULATIONS (AT MINIMUM RATES)

------------------------------------------------

FRACTIONAL WEEPAGE ( 0.20 MAXIMUM) 0.032

$ DOWNCOMER SEAL, INCHES (-0.25 MINIMUM) 0.310/ 0.572

--------------------------

DESIGN LIQUID RATE (L) GPM/INCH OF WEIR/PASS 10.487/ 6.861

VAPOR LOAD/FREE AREA FT/SEC 0.170/ 0.181

JET FLOOD (VL/AF) ALLOW FT/SEC 0.199/ 0.238

SURFACE TENSION - VISCOSITY PARAMETER 0.933

MAXIMUM RECYCLED VAPOR, % 0./ 0.

TRAY FROTH DENSITY (OUT/INBOARD) 0.252/ 0.243

(FRACT FROTH VOL OCCUPIED BY LIQ)

EST. LIQUID HOLDUP (DECK+DC), FT3 21.415/ 15.855

EST. DOWNCOMER LIQ. HOLDUP, FT3 9.813/ 5.613

* DENOTES INPUTTED HARDWARE INFORMATION

++ LOW/MODERATE PRESSURE CORRELATION USED FOR MAX INLET VELOCITY

$ WEEP CORRECTED CLEAR LIQUID HEIGHT NOT USED IN CALCS

## INBOARD PASS CHORD LENGTH USED FOR INBOARD PASS CLEAR LIQUID

HEIGHT CALC AS OF NOV. 1998

DESIGN PRACTICES

SIEVE TRAY EFFICIENCY CALCULATIONS - OUTBOARD PASS

VERS. 7.5

COPY NUMBER 1

ALL CALCULATIONS ON THIS PAGE ARE MADE AT DESIGN RATES AND

INCLUDE THE EFFECTS OF WEEPING EXCEPT WHERE OTHERWISE NOTED

EQUILIBRIUM PARAMETERS

----------------------

COMPONENT EQUILIBRIUM LAMBDA

KEY COMP NO.1 0.700 0.411

MASS TRANSFER PARAMETERS

------------------------

VAPOR MASS TRANSFER COEFFICIENT, KG CM/SEC 0.887

LIQUID MASS TRANSFER COEFFICIENT, KL CM/SEC 0.096

NG 1.763

NL 4.368

COMPONENT NOG PERCENT LIQUID

PHASE CONTROL

KEY COMP NO.1 1.512 14.2

PHYSICAL PROPERTIES AND LOADINGS

--------------------------------

LIQUID RATE LB-MOLES/HR 4454.382

LIQUID MOLECULAR WEIGHT LB/MOLE 125.903

LIQUID MOLECULAR DIFF (FRI) CM2/SEC 0.745E-04

VAPOR RATE LB-MOLES/HR 2616.581

VAPOR MOLECULAR WEIGHT LB/MOLE 123.283

RESIDENCE TIME CALCULATIONS

---------------------------

FRACTION WEEPING 0.002

CLEAR LIQUID HEIGHT INCHES 1.883

FROTH DENSITY 0.252

LIQUID RESIDENCE TIME SECONDS 2.880

VAPOR RESIDENCE TIME SECONDS 0.499

--------------------------

EFFECTIVE FLOW PATH LENGTH FT 3.802

NUMBER OF MIXING POOLS 32.180

INTERFACIAL AREA CM2/CM3 3.982

DESIGN PRACTICES

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

* DESIGN NO WEEP MIN. RATE *

* POINT TRAY OVERALL OVERALL OVERALL *

* EFFIC. EFFIC. EFFIC. EFFIC. EFFIC. *

* KEY COMP NO.1 70.2 80.4 72.1 72.1 68.9 *

* * * * * ALL EFFICIENCIES DEBITTED 10% ON POINT EFFICIENCY * * * * * *

Normal Program Completion

DP03B

design practices

downcomer clearances

downcomer width

downcomer layout

global practices

hole area

tray balancing

basic design considerations

Documents