Tài liệu về tháp chưng cất

the separa-tion efficiency, the pressure drop, the liquid holdup, the capacity, and the costs.A physical model is required to describe the hydraulics and the mass transfer efficiency in

The clamour for energy-saving techniques in almost all branches of industry has acted as

a spur in the development of thermal separation equipment The design and process neering improvements that have ensued entail that feedstocks are subjected to less severetreatment and can thus be optimally exploited They also entail production under ecologi-cally favourable conditions (cf Fig 1.1)

engi-A typical example is provided by low-pressure-drop packing in the vacuum rectification ofmixtures that are unstable to heat and that necessitate a large number of theoretical stagesfor their thermal separation The attendant decrease in the total pressure drop and operationunder vacuum ensure that the temperature at the bottom of the column is comparativelylow Hence, decomposition products that are detrimental to the environment can be largelyavoided, i.e atmospheric pollution is reduced and less residues have to be disposed of.Another advantage is that the reduction in the average column pressure brought about byvacuum operation increases the average relative volatility of the components in the mixtureand thus reduces energy consumption

Low-pressure-drop, high-performance packing is an essential requirement in the economicdesign of an integrated separation plant, because it permits heat pumps to be installed and anumber of columns to be linked together

Fig 1.1 Relationships established by separation techniques between energy consumption, sing and environmental protection

proces-Packed Towers in Processing and Environmental Technology Reinhard Billet

Copyright © 1995 VCH Verlagsgesellschaft mbH, Weinheim

ISBN: 3-527-28616-0

From this point of view, it is not surprising that packing designed to conserve energy hasbeen the subject of many new developments on the part of equipment manufacturers How-ever, before any one particular type of packing can be selected for a given separation task,adequate knowledge must be available on the performance characteristics, e g the separa-tion efficiency, the pressure drop, the liquid holdup, the capacity, and the costs.

A physical model is required to describe the hydraulics and the mass transfer efficiency in

a given separation column and thus to allow the main dimensions to be calculated and theprocess engineering performance to be predicted The parameters that affect the design, thecapacity, and the specific properties of the product and inlet streams must be known in order

to devise the model, and the only means of acquiring these data is by experiment

Consequently, the aim of this book is to supplement the theoretical considerations by theresults of relevant studies that were performed in the author's laboratories and pilot plantsand were scaled up to meet practical requirements It is known that packed columns,whether random or stacked, allow lower pressure drops per theoretical stage than platecolumns and are thus better suited to meet demands on optimum energy consumption inthermal separation plants

The main applications for packed columns are the separation of vapour-liquid orgas-liquid systems, e.g in rectification, absorption, and desorption In many cases, theyhave also proved to be superior to conventional plate columns for liquid-liquid extractionprocesses Greatest significance from the aspect of saving energy is attached to their use inrectification, a subject to which particular attention has been devoted in the course of thisbook

In future, these separation processes will not be confined to petroleum refineries and thechemical and allied industries They will also be adopted on a wider scale in ecological engi-neering for purifying off-gas streams and for water treatment, and the demand for the neces-sary equipment, including packing, will grow accordingly

Rectification, absorption, desorption, and liquid-liquid extraction processes consist tially of passing two countercurrent phases through a packed column (cf Fig 1.2)

essen-In rectification, the vapour produced in the distillation section of the column flows tercurrent to the liquid formed in a condenser Contact between the two phases is thusestablished, with the result that the low-boiling component flows upwards and the high boileraccumulates at the bottom of the column

coun-Physical absorption processes consist of mass transfer from the gas into the liquid phase,

i e into a solvent that selectively absorbs the desired component from the gas stream Indesorption - often referred to as stripping - mass transfer proceeds in the opposite direction,i.e from the liquid into the gas phase

In the extraction of a component from a mixture of liquids by means of a selective vent, mass transfer takes place between two liquid phases

sol-It is taken for granted here that the reader is already acquainted with these separationprocesses and their thermodynamic fundamentals together with the literature on the subjectand the standard terminology The scope of this book has therefore been restricted to a com-prehensive treatment of packed-bed technology and its application to separation processes

Packed towers for continuousRectification Absorption Desorption Extraction

Solvent I A Liquid = A Loaded(absorbent)J { b a s feed : | gas

Fig 1.2 Applications for packed columns in thermal separation processes

1.1 Fundamental operating characteristics of packed columns for gas-liquid systems

The optimum choice of a packed column for a given separation task would be impossiblewithout a sound knowledge of the characteristic parameters and physical quantities that per-mit process engineering evaluation and comparison

The gas or vapour load in a packed column is usually expressed by the capacity factor,which is given by

where

uv is the superficial velocity,

o v is the density of the gas or vapour, and

QL is the density of the liquid

At low or moderate pressures, Q V is small compared to Q L and can thus be neglected.Hence, the vapour capacity factor, which is often resorted to in practice as a measure of the

dynamic load on the column, is the product of the superficial velocity u v and the square root

of the vapour density, i.e

(1-2)

The maximum permissible value for the capacity factor also depends on a dimensionless

flow parameter that allows for the ratio of the liquid to the vapour flow rate LIV and is

(1-5)

Another factor of great importance in evaluating packed columns is the pressure drop per

unit of separation efficiency, which is defined by the number of theoretical stages n t in

recti-fication or by the number of transfer units NTU in absorption and desorption A general term, i.e the number of separation units N required for a given process, may be taken to

cover both cases Thus,

Liquid Gas or vapour

N = nt or N = NTU (1-6)

The total pressure drop in the gas or vapour is given by

&P=PB-PT (1-7)

where p B is the pressure at the lower column inlet and

pT is the pressure at the upper column outlet

In analogy to Eqn (1-5), the total pressure drop per separation unit can be expressed as a

function of the capacity factor for a given ratio LIV of the liquid to the vapour flow rate,

i.e

~ - = HFy) (1-8)

The following relationship thus applies:

^ (1-9)

If the vapour pressure at the top of the column is kept constant during rectification, that

at the bottom of the column - and thus the boiling point - become less as the pressure dropper unit efficiency for the specific bed of packing decreases Low boiling points are an abso-lute necessity in the vacuum rectification of thermally instable mixtures They also entaillarger differences between the temperature of the heating medium and that of the mixture,i.e more effective heat transfer Moreover, if a large number of theoretical stages isrequired in a given vacuum rectification, a low pressure drop allows entire installations to bedesigned from the aspect of optimum energy consumption

The number of theoretical stages n t required to separate a binary feed with a molar flow

rate F can be determined diagrammatically by the McCabe-Thiele method, which is based

on the concept of an equilibrium stage, i.e a stage in which the ascending vapour is inphase equilibrium with the descending liquid An example of the corresponding diagram is

shown in Fig 1.4, in which x ¥ is the mole fraction in the feed of the more volatile

compo-nent that has to be concentrated to a mole fraction x D in the overhead product (distillate)

with a molar flow rate D; and x B is the mole fraction of this volatile fraction that remains in

the bottom product (molar flow rate B) The equilibrium curve is the locus of the points x,

y The two operating lines derived from material balance equations - BI for the stripping zone and DI for the enrichment zone - intersect at the point / on the g-line, which is the

locus of all points of intersection of the two operating lines for any given feed stream

The number of theoretical stages n t is represented by the steps that link the equilibriumcurve with the two operating lines between the points xB and xD It is the sum of the number

of stages in the enrichment zone n tez and the number in the stripping zone n tsz, i e.

Liquid mole fraction x

Fig 1.4 Determination of the the number of theoretical stages in a continuous rectificationcolumn, taking a binary mixture as an example

If x F, xD and x B are given, n t depends on the phase equilibrium in the mixture, i e onthe shape of the equilibrium curve for the binary system, and on the position of the two

operating lines BI and DI.

The position of the intersect / can be obtained from the material balance at the inletcross-section of the column, i.e

The factor / in this equation is a measure for the thermal state of the feed and is given by

h' F -h F

/ = ! - "

where h F is the molar enthalpy of the feed at the inlet temperature, h' F is the molar enthalpy

at the operating temperature in the inlet cross-section of the column, and Ah v is the molarcondensation enthalpy of the vapour stream in the inlet cross-section

If the individual components of the mixture have roughly the same molar evaporationenthalpy, the molar flow rates of both the vapour and liquid will remain practically constantalong the respective flow paths both above and below the inlet, i.e in both the enrichment

and stripping zones In this case, the relationship between the vapour y and the liquid x

mole fractions of the more volatile component in the enrichment zone is given by the ing linear equation:

follow-y = r + 1 • x +

r + 1 (1-13)

where r is the ratio of the reflux flow rate L to the overhead product flow rate D, i.e.

r = | - d-14)

The corresponding equation for the stripping zone is

where b is the ratio of the flow rate of liquid L in the stripping zone to that of the bottom product B, i e.

b = -j (1-16)

If the values for xF and / are given and those for xD and xB are specified, the operating

lines for the enrichment and stripping zones in Fig 1.4 can be easily plotted Inserting x = 0

in Eqn (1-13) yields the value for y0 at the intercept with the axis of ordinates, i.e.

d-17)

where xD is the mole fraction of the more volatile component in the overhead product.

The reflux ratio r at the head of the column is a factor that greatly affects the economics

— (1-20)

Whether its position is higher or lower than the corresponding intercept formed by thelinear relationship {cf Eqn (1-17)} depends on which of the two components - the high-boiler or the low-boiler - has the greater molar evaporation enthalpy.

Likewise, the relationship for the stripping zone in the column is also nonlinear if themolar evaporation enthalpies of the components differ In this case, the following equationapplies:

and y t and x t are the coordinates of the intersect / (cf Fig 1.4)

An alternative concept to the number of theoretical stages n t for the evaluation of

separa-tion efficiency is the number of transfer units NTU If the molar flow rates for the liquid L and the vapour V are kept constant and there is no mixing in the axial direction, the NTU

for steady-state operation can be expressed as follows in the terms of the concentration

dif-ference y* - y in the vapour (cf Fig 1.5):

—£- (1.24)

where y* is the phase equilibrium concentration of the more volatile component in the

vapour in contact with the liquid of concentration x* at the phase boundary in anygiven horizontal cross-section of the column; and

x and y are the fractions that correspond to the concentrations of the more volatile

component in the bulk of the vapour and in the liquid respectively

It may often be assumed that the resistance to mass transfer in rectification is

predomi-nantly in the vapour phase, i e x —» x* In this case, the surface mass transfer coefficient in

the liquid phase is |3L —» °o and that in the vapour phase fV is identical to the overall mass

transfer coefficient on the vapour side k ov Accordingly, NTUV = NTUOV The height of a

transfer unit is then defined by

H H HTUOV =

where a ph is the phase contact area per unit column volume,

ds is the column diameter,

V is the molar flow rate of the vapour, and

HTUOV is usually determined by experiment

Hence the following relationship exists between the number of theoretical stages per unit

height n t/H and the height of a transfer unit HTUOV:

In other words, the number of transfer units in systems with a low relative volatility isgiven by

(NTUov)a = smal, = n, (1-28)

Likewise, the height of a transfer unit can be equated to the height equivalent of a retical stage, i.e

theo-{HTUOy) a , small = HETS = — (1-29)

The stripping factor X for a given section of the column is defined as the product of the mean slope m yx of the equilibrium curve and the molar vapour/liquid ratio VIL, i e.

\ = myxj- (1-30)

Once the height of a transfer unit HTU OV and the number of transfer units NTU OV areknown, the height of the column required for the relevant separation process can beobtained from their product Hence,

c y

H = HTUQV ' NTUOV = HTUOV ^— (1-31)

The figure thus derived is identical to that obtained from the number of theoretical stages

nt and the separation efficiency n tIH for the packing concerned {cf Eqn (1-27)}, i e.

H =

An analogous analysis applies for mass transfer in the liquid phase

This procedure for the determination of column height is referred to in the literature asthe HTU-NTU concept It is applied in Chapters 3, 5 and 14

The efficiency, expressed as the number of separation units per unit height NIH, can be

assessed graphically from its relationship to the column load or capacity factor, which can bewritten as

NIH = i{Fv) (1-33)

A knowledge of this function is essential in determining the load relationships for the

pressure drop per unit height AplH {Eqn (1-5)} and the pressure drop per separation unit

vv=i(Fv) (1-34) vv=f(Ap/N) (1-35) The specific column volume v v is an important factor in determining the capital invest-ment costs and is defined by

where H is the effective height of the column,

N is the number of theoretical separation units, and

uv is the superficial vapour velocity

1.2 Theoretical column efficiency

The relative volatility a, a term that was introduced in Eqns (1-28) and (1-29), is a sure for the ease with which a mixture can be separated It expresses the relationship

mea-between the molar fraction y of the more volatile component in the vapour and that of the liquid x with which it is in phase equilibrium; and it is defined by

Its magnitude governs the shape of the equilibrium curve (cf Fig 1.4), which is defined

by the following equation:

If a is constant, the curve assumes the form of an equilateral hyperbola The greater the

value of a, the greater the value of y for a given value of x.

As a rule, the relative volatility a of the two components of a mixture, and thus the easewith which they can be separated, decreases with a rise in pressure Thus, if the value of a

for a given mixture is comparatively small, the number of theoretical stages n t required for

separation increases with the specific pressure drop Ap/n t within the packing, particularly

during vacuum operation If the pressure at the head of the column p T is kept constant, the

pressure drop Ap governs the rise in pressure up to the value at the bottom of the column

pB, i.e.

[^y (1-39)

It follows that the reflux ratio required for separation in the enrichment zone of a cation column rises In other words, the energy consumption must increase In view of thefollowing relationship, the liquid-vapour ratio also increases: