close

Вход

Забыли?

вход по аккаунту

?

Microwave regeneration of adsorbents for VOCs removal

код для вставкиСкачать
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the
text directly from the original or copy submitted.
Thus, som e thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
T he quality of th is reproduction is d ep en d e n t upon th e quality o f th e co p y
subm itted.
Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleedthrough, substandard margins, and improper alignment
can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and
there are missing pages, these will be noted. Also, if unauthorized copyright
material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning
the original, beginning at the upper left-hand comer and continuing from left to
right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6° x 9” black and white photographic
prints are available for any photographs or illustrations appearing in this copy for
an additional charge. Contact UMI directly to order.
Bell & Howell Information and Learning
300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA
UMI
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
M icrowave Regeneration of Adsorbents for VOCs Removal
by
Maria I. Hemindez Rodriguez
A project report submitted in partial
fulfillment of the requirements for the
degree of
Master of Engineering
in
Chemical Engineering
University of Puerto Rico
Mayagiiez Campus
1999
Approved by:
Abraham Rodriguez-Kamirez, Ph.'D.
Member, Graduate
1
Dati
^ 0 -4 -—
Gilberto Villafarie-Ruiz, Ph. D.
Member, Graduate Committee
1
.
Date ^
e Benitez
ent, Graduate C
/ |n
Sanluel Hemindez, Ph. D.
Representative of Graduate School
^
u Xm a l o
Federico Padron
Chairperson of the Department
Maria Aponte
Director, Graduate School, Ph. D.
I
I.
** I
^
! ^
J
Dat^y
AuW
2 .
1 ____
Date^
T M
Date
H
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1________
UMI Number 1396675
___
®
UMI
UMI Microform 1396675
Copyright 2000 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract
This study addresses the technical and economical feasibility of a full-scale,
microwave-regenerated (MW) volatile organic compound (VOC) adsorption plant It
compares the estimated economics of MW-regenerated adsorption systems with those of
other conventional VOCs control systems. The VOC emission stream chosen as the basis
for the study is typical of a large-scale industrial printing operation. The concentration
and flow rate are representative of an enclosed plant in which the VOCs are emitted and
collected in a single stream (flow conditions: 68 m3/s at 500 ppmv). The VOC is assumed
to be methyl ethyl ketone (MEK), a common water-soluble industrial solvent with
representative dielectric and adsorption properties. The adsorbent chosen was Dowex
Optipore, a hydrophobic polymeric adsorbent produced by Dow Chemical.
It was found that adsorption of MEK on Dowex Optipore at ambient temperature is
described by the Langmuir adsorption isotherm model for gas phase concentrations as
high as 20,000 ppmv. The best values of the corresponding parameters, found by leastsquares estimation after linearizing the model, were 3.964 kg MEK/kg Optipore-kPa and
10.083 kPa"1. The use of the Thomas solution for breakthrough calculations during the
adsorption and cooling phases of the cycle greatly simplified the calculations when
compared to the Michael graphic method.
For this case study, a MW-regenerated adsorption system consisting of 13 beds, 4.2 m
in diameter and 0 3 0 4 m in depth, each one loaded with 1,440 kg of Optipore, results in a
net profit of $0.039/lb of MEK recovered.
i
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Resumen
Este estudio evalua la viabilidad tdcnica y econdmica de una planta de adsorcion
de compuestos organicos volatiles (VOCs por sus siglas en ingles) a escala industrial,
regenerada usando microondas (MW por sus siglas en ingles). Compara los aspectos
econdmicos de este proceso con otros sistemas convencionales de control de emisiones
de VOCs. El caso de estudio de emisiones de VOC seleccionado es tfpico de una
imprenta a gran escala. Los flujos y concentraci ones son representatives de una planta
encerrada en la que todas las emisiones de VOCs son capturadas en una sola corriente
(condiciones de flujo: 68 m3/s a 500 ppmv). Se presume que el VOC es metil etil cetona
(MEK por sus siglas en inglds), un solvente industrial muy comun que es soluble en agua,
con propiedades dieldctricas y de adsorcidn representativas. El adsorbente seleccionado
es Dowex Optipore, un sdlido polimdrico, hidrofdbico producido por Dow Chemical.
Se encontrd que la adsorci6n de MEK en Dowex Optipore a condiciones
ambientales se puede describir usando el modelo de isoterma de Langmuir hasta
concentraciones del gas tan altas como 20,000 ppmv. Los mejores estimados de los
parametros, evaluados por el metodo de cuadrados mmimos luego de linearizar el
modelo, son 3.964 kg MEK/kg Optipore-kPa y 10.083 kPa'1. Se utilizo la solucidn de
T hom as para predecir el comportamiento dindmico del lecho estdtico durante adsorcion y
enfriamiento, lo que simplified los cdmputos.
Para este caso de estudio, un sistema de adsorcion regenerado por MW que
consiste de 13 lechos de 4.2 m de didmetro y 0304 m de profundidad, cada uno cargado
con 1,440 kg de Optipore, resulta en una ganancia neta de $0.039/lb de MEK recuperado.
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To Mami, Papi, Evelyn, and Vivian for always being there...
To titi Evelyn who is my inspiration...
To Abuela Chencha with her strong sp irit...
To Benitez for teaching me a lot...
To God, who gave me all of them, all and much more.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgments
The completion of this project would not be possible without the following:
•
Dr. Jaime Benitez-Rodrfguez, President of my Graduate Committee;
•
Dr. Abraham Rodriguez and Dr. Gilberto Villafane, Members of my
Graduate Committee.
•
University of Puerto Rico, Mayagiiez Campus, Department of
Chemical Engineering Faculty and Personnel
•
Dow Chemical Technical Support Department
•
Dr. Price and Dr. Schmidt from the University of Texas at Austin
•
Benitez’s family and friends for their support.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table o f Contents
Chapter 1. Introduction............................................................................................................1
1.2 Objective......................................................................................................................3
Chapter 2. Theory................................................................................................................... 4
2.1 Introduction to gas adsorption as an air pollution control technology.....................4
2.2 Adsorption equilibria................................................................................................... 7
2 3 Dynamics of fixed-bed adsorption............................................................................ 11
23.1 Method of Michaels (1952)................................................................................ 13
2 3 .2 The Thomas solution.......................................................................................... 20
2.4 Heat and momentum transfer in fixed-bed adsorbers.............................................. 22
2.4.1 Cooling of a thermally regenerated fixed-bed................................................. .22
2 .4 3 Pressure drop across fixed-beds..........................................................................24
2 3 Regeneration of fixed-bed adsorbers using microwave radiation. ........................ 24
23.1 Interaction of microwaves with polar/polarizable m aterials........................... 26
Chapter 3. Previous Work..........................................................
29
Chapter 4. Methodology......................................................... - ...........................................3 3
Chapter 5. Results................................................................................................................ 3 5
5.1 Breakthrough calculations using Thomas method....................................................3 5
5.2 Effect of temperature on adsorption equilibrium......................................................3 9
5 3 Cooling of the thermally regenerated bed............................................................... 41
5.4 Pressure drop across the bed..................................................................................... 42
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5.5 Design considerations for fixed beds......................................................................... 42
5.6 Cost for the Dowex Optipore adsorption system....................................................... 43
Chapter 6. Discussion of Results...................................................................................
48
Chapter 7. Conclusions...............................................................
51
References......................................................... - ....................................................................52
Appendix A. FORTRAN Computer Program for J Function................
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
L ist of Figures
Figure 1. Schematic diagram of a fixed-bed adsorption system (Benitez, 1993).................6
Figure 2. Equilibrium adsorption of vapors on activated carbon (Treybal, 1980)............... 7
Figure 3. Typical breakthrough curve...................................................................................13
Figure 4. The adsorption zone................................................................................................18
Figure 5. Modeling organization for microwave (MW) regeneration studies.................. 34Figure 6. Langmuir isotherm of MEK on Dowex Optipore (25 ° Q .................................35
Figure 7. Adsorption of MEK on Dowex Optipore, effect of temperature.......................40
Figure 7. Regeneration kinetics at equilibrum pressure o f 3 3 3 kPa for MEK/Dowex
O ptipore.................................................................................................................................. 41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
Table 1. Typical Physical and Chemical Properties of Dowex Optipore........................ 3 6
Table 2. Feed and Equilibrium Conditions Through the Bed for 500 ppm of MEK in Air
at 298 K and 1013 k P a..........................................................................................3 6
Table 3. Mass Transfer Coefficients for Mixture o f 500 ppm of MEK in A ir at 298 K and
1013 kPa................................................................................................................. 3 8
Table 4. Regeneration of M EK from Dowex Optipore at 3 3 3 kP a.................................3 9
Table 5. Average Cost Factors for Adsorbers.......................
44
Table 6. Factors and Values for Estimating Adsorbers Annual C osts............................. 46
Table 7. Control Technologies for the Treatment of MEK at 500 ppm in 67.96 m3/s
Studied by Price and Schmidt (1998b)....................................................................47
Table 8. Common Industrial Solvents which Favor M W Regeneration (Price and
Schmidt, 1998b)...............................................................................
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
Chapter 1. Introduction
Regulation of the emission of volatile organic compounds (VOCs) from industrial
sources is a major element of the Clean Air Act Amendments of 1990. In the next decade,
industries using organic solvents will make major investments in VOC abatement; these
include printing, metal fabricating, automotive, and semiconductor manufacturing among
others. Adsorption is one of the most common methods for removing low concentrations
of VOCs from gaseous emission streams. To maintain the adsorbent's ability to capture
VOCs it must be periodically regenerated. This typically entails stripping the organics
from the bed with a hot inert gas or with steam. While the concentration of VOCs in the
resulting purge gas is significantly greater than in the original process stream, recovery of
these compounds is generally not economical due to the still very low dew point
temperatures and/or the need to dispose of hazardous wastewater in the case of watersoluble VOCs. In most cases, the VOCs are simply incinerated.
Recovery of VOCs from industrial emission streams offers inherent advantages
over destructive means of abatement (i.e., incineration) from both the regulatory and user
perspectives. These include the elimination of combustion product emissions and use of
nonrenewable natural resources to incinerate the VOCs, as well as elimination of
emissions and resource consumption required to create the replacement feedstocks to the
industrial operation. From the user perspective, a costly environmental problem stream
may be converted to a valuable resource stream. For certain VOCs, microwave
regenerated adsorption systems may offer a cost-effective means of recovery for reuse.
1
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
2
As an alternative to destructive abatement methods, it may be possible to recover
VOCs by using dielectric heating to regenerate the adsorbent bed. While convective
heating requires the use of a heating medium (the stripping gas), microwave radiation can
generate heat directly within the adsorbent bed. Thus, regeneration can take place under
vacuum or with a much reduced volume of stripping gas; VOCs can then be easily
recovered from the non-diluted regeneration off-gas stream by direct condensation using
near-ambient temperature cooling water.
Microwave heating has seen extensive use in industrial drying due to the
advantages associated with selective volumetric heating of water. Several researchers
have extended these ideas to the regeneration of adsorbents (Price and Schmidt, 1997;
Gibson, et al., 1988; Burkholder, et al., 1986; Roussy, et al., 1984). Previous experiments
have shown that very high heat-transfer rates to the adsorbent bed can be achieved,
limited only by the capacity of the microwave generator. Moreover, mass transport out
of the adsorbent is enhanced by a pressure-driven flow and the mass-transfer rates are so
high that desorption follows a quasi-equilibrium process.
By contrast, conventional
regeneration typically is rate-limited by heat and mass transfer resistances, resulting in
longer regeneration times.
This study addresses the technical and economical feasibility of a full-scale
microwave regeneration plant It compares the estimated economics of microwave
regeneration systems with those of other conventional VOCs control systems. The VOC
emission stream chosen as the basis for the study is typical of a large-scale industrial
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
printing operation, an important area of application. The concentration and flow rate are
representative of a plant in which the VOCs are emitted into and collected from the
building permanent total enclosure (PTE flow conditions: 68 m3/s at 500 ppmv). The
VOC is assumed to be methyl ethyl ketone (MEK), a common water-soluble industrial
solvent with representative dielectric and adsorption properties.
IJ2 Objective
The objective of this work is to determine the optimum operational conditions of
a microwave regenerated VOC adsorber using Dowex Optipore to adsorb methyl ethyl
ketone (MEK) from the emissions of an industrial printing operation. Another objective
is to compare the economics of this technique with alternate techniques for VOCs
abatement
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 2. Theory
2.1 Introduction to gas adsorption as an air pollution control technology
Gas adsorption, is a separation process based on the ability of certain solids to
remove gaseous components preferentially from a flow stream. In air pollution control
applications, the pollutant gas or vapor molecules present in a waste gas stream collect on
the surface of the solid material. Adsorption is useful in removing objectionable odors
and pollutants from industrial gases as well as recovering valuable solvent vapors from
air and other gases. It is a particularly useful technique when the gaseous emissions
present in the waste gas are valuable enough to recover for recycling or resale.
The components and operation of a typical fixed-bed adsorber system as shown
on Figure 1 will be described. The system shown in that figure uses two vessels in which
beds of adsorbent are located. The waste gas from a process enters the main blower and
passes through a cooler. The reason for cooling is that the VOC-adsorbing capacity of the
solid increases as the temperature decreases since adsorption is an exothermic process.
The cooled gas stream passes through one of the adsorbent beds, where most of the VOC
is removed. The “clean air” is either vented to the atmosphere or returned to the source
process. Eventually, a substantial portion of the bed becomes saturated with the VOC,
and the pollutant starts to “breakthrough” in the effluent. When that happens, the waste
gas stream is switched to the idle bed, which has been regenerated and cooled.
The expended bed is then regenerated by direct contact with steam. The adsorbed
VOC is displaced from the adsorbent, and the mixture of steam and organic vapor is
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
condensed and collected in a decanter for initial separation. If the solubility o f the
condensed VOC in water is low enough, decanting is sufficient; otherwise an additional
separation unit, such as distillation, is required. The VOC is recovered for reuse or sale,
and the waste water, usually, must be further purified for reuse or discharge.
The adsorbents most frequently used for air pollution control include activated
carbon, alumina, bauxite, and silica gel. Activated carbon is, by far, the m ost frequently
used adsorbent, and has virtually displaced all other materials in solvent recovery
systems.
The term activated as applied to adsorbent materials refers to the increased
internal and external surface area imparted by special treatment processes. Any
carbonaceous material can be converted to activated carbon. Coconut shells, bones,
wood, coal, petroleum coke, lignin, and lignite all serve as raw materials for activated
carbon. However, most industrial grade carbon is made from bituminous coal (Cooper
and Alley 1986).
Activated carbon is manufactured by first dehydrating and carbonizing the raw
material. Activation is completed during a controlled oxidation step in which the
carbonized material is heated in the presence of an oxidizing gas. For certain carbons the
dehydration can be accomplished by using chemical agents. The ideal raw material has a
porous structure that provides a uniform pore distribution and high adsorptive capacity
when activated. Activated carbon is tailored for a specific end use by both raw material
selection and control of the activation process. Carbons for gas-phase applications have a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
specific surface area in the range of 800 to 1,200 m2/g, and a porosity in the range of 35%
to 40%. Most of the pore volume is distributed over a narrow range of pore diameters,
usually ranging from 0.4 to 3.0 nm. Bulk density is of the order of 500 kg/m3.
Vents
Steam
Condenser
Decanter
Adsorbers
Steam-VOC
mixture
VOC pump
( Cooler)
Aqueous layer
Main blower
Air-VOC mixture
Waste pump
Figure 1. Schem atic d iag ram of a fixed-bed adsorption system (Benitez, 1993).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
2J2 Adsorption equilibria
In many aspects the equilibrium adsorption characteristics of a vapor or gas on a
solid resemble the equilibrium solubility of a gas in a liquid (Treybal, 1980). Figure 2
shows several equilibrium adsorption isotherms for a particular activated carbon as
adsorbent. The concentration of adsorbed gas (adsorbate) on the solid is plotted against
the equilibrium partial pressure, p*, of the vapor or gas at constant temperature.
Examples of such isotherms are shown on Figure 2. At 373 K, for example, pure acetone
vapor at a partial pressure o f 2533 kPa is in equilibrium with an adsorbate concentration
of 0.2 kg adsorbed acetone/kg carbon. Increasing the partial pressure o f acetone will
cause more to be adsorbed, and decreasing it will cause acetone to desorb from the
carbon.
25-
15-
10Acetme,
303 K
0
005 01 0.15 0.2 025 03
Capacity, kg adscibedAg carbon.
035 0.4
F igure 2. E quilibrium adsorption o f vapors on activated carbon (T reybal, 1980).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
Because adsorption is an exothermic process, the concentration of adsorbed gas
decreases with increased temperature at a given equilibrium pressure, as the several
acetone isotherms of Figure 2 illustrate. Different gases and vapors are adsorbed to
different extents under comparable conditions. Figure 2 shows that benzene is more
readily adsorbed than acetone at the same temperature and gives a higher adsorbate
concentration for a given equilibrium partial pressure. As a general rule, gases and vapors
are more readily adsorbed the higher their molecular weight The adsorption equilibrium
is highly dependent on the solid used as adsorbent For example, the equilibrium curves
for acetone and benzene on silica gel would be entirely different from those of Figure 2.
Even differences in the origin and method of preparation of a given adsorbent will result
in significant differences in the equilibrium adsorption.
There are two distinct adsorption mechanisms: physical adsorption and
chemisorption. Physical adsorption, also referred to as van der Waals adsorption,
involves a weak bonding of gas molecules to the solid. The bond energy is similar to the
attraction forces between molecules in a liquid. The adsorption process is exothermic,
and the heat of adsorption is slightly higher than the heat of vaporization of the adsorbed
material. The forces holding the gas molecules to the solid are easily overcome by either
the application of heat or the reduction of pressure. Most of adsorption applications to air
pollution control involve physical adsorption.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
Chemisorption involves an actual chemical bonding by reaction o f the adsorbate
with the adsorbing solid. It is virtually irreversible. Except in some very specialized
applications, recovery of a substance through chemisorption is not feasible.
One of the most useful mathematical models to describe adsorption equilibria is
the Langmuir isotherm. Its theoretical development is based on the following
assumptions: (1) the adsorbed phase is a unimolecular layer, and (2) at equilibrium, the
rate of adsorption is equal to the rate of desorption from the surface. Define / as the
fraction of the total solid surface occupied by adsorbate molecules. The rate of
adsorption. ra, is proportional to the partial pressure of the adsorbate, p , and to the
fraction of the solid surface area available for adsorption, (1 —f). Therefore,
= Cmp ( l - f )
(2-1)
where Ca is a constant Conversely, the rate of desorption, rd, is proportional to the
fraction o f the surface area occupied by the adsorbate:
(2.2)
where Cd is a constant At equilibrium, the rate of adsorption is equal to the rate of
desorption. The fraction of the surface covered is, then, given by
(23)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
Because the adsorbed phase is a unimoiecular layer, the mass of adsorbate per
unit mass of adsorbent, m, is also proportional to the surface covered:
m = C j,,/
(2*4)
where Cmis a constant. Combining Eqs. (23) and (2.4),
« - - JlsL.
(zs)
&2P* + 1
where kf = CaC JC d , and fcj = CJCd. Equation (23) is known as Langmuir isotherm. At
very low adsorbate equilibrium partial pressure hp* is approximately equal to zero, and
Eq. (2.5) becomes
^ p*
(2.6)
Conversely, at high equilibrium partial pressure,
A
(2*7)
Hence, over an intermediate range of partial pressures:
m * k{p * )n
(2-8)
where:
k = constant
n = constant with a value between 0 and 1
Equation (2.8) is known as the Freundlich isotherm. Even though it is not based on a
rigorous theoretical background, the Freundlich isotherm gives an adequate description of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
adsorption equilibrium in many air pollution control applications. The values of k and n
are obtained from experimental data.
2 3 Dynamics o f fixed-bed adsorption
Steady-state adsorption requires continuous movement of both fluid and
adsorbent through the equipment at constant rate, with no change in composition at any
point in the system with passage of time. The inconvenience and relatively high cost of
continuously transporting solid particles as required for steady-state operation makes it
more economical to pass the fluid mixture to be treated through a stationary bed of
adsorbent As increasing amounts o f fluid pass through such a fixed-bed, the solid
adsorbs increasing amounts of adsorbate, and an unsteady state prevails. The dynamic
behavior of such an operation is the subject of this section.
Consider a binary gaseous mixture with an adsorbate concentration o f Cq. The gas
passes continuously down through a relatively deep fixed-bed of adsorbent initially free
of adsorbate. The uppermost layer of solid at first adsorbs rapidly and effectively, and
what little adsorbate is left in the solution is substantially all removed by subsequent
layers of solid in the lower part of the bed. The bulk of the adsorption occurs over a
relatively narrow adsorption zone in which the concentration changes rapidly. The
effluent from the bottom of the bed is practically adsorbate free. As the gas continues to
flow, the uppermost layer of the bed becomes saturated, and the adsorption zone moves
downward as a wave at a rate ordinarily much slower than the linear velocity of the fluid
through the packed bed. When the lower portion of the adsorption zone reaches the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
bottom of the bed, the concentration of adsorbate suddenly rises to an appreciable value
for the first time. The system is said to have reached the breakpoint. The adsorbate
concentration in the effluent now rises rapidly as the adsorption zone passes through the
bottom of the bed. Soon the bed is completely exhausted and the outlet composition is
exactly equal to the inlet composition. The portion of the effluent concentration curve
between the breakpoint and exhaustion is termed the breakthrough curve. Figure 3 shows
a typical breakthrough curve.
The shape and time of appearance of the breakthrough curve greatly influence the
design and the method of operating a fixed-bed adsorber. The curves generally have an S
shape, but they may be steep or relatively flat The breakpoint is very sharply defined in
some cases and in others poorly defined. The rate of the adsorption process, the nature of
the adsorption equilibrium, the fluid velocity, the feed concentration, and the length of
the bed all contribute to the shape of the curve for a given system.
The effective rate of adsorption is determined by one or more of several
diffusional steps. Individual steps in the transport mechanism follow.
•
Diffusion in the sorbed state (in a uniform liquid-like or solid phase or a pore-suxface
layer). This is designated as particle-phase diffusion.
•
Reaction at the phase boundary, usually very fast
•
Pore diffusion in the fluid phase, within the particles.
•
Mass transfer from the fluid phase to the external surfaces of the adsorbent particles.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
•
Mixing, or lack of it, between different parts of the contacting equipment. For
instance, in column operation with slow flow rates, the breakthrough curve may be
broadened by axial dispersion.
Inlet, Cq
Adsorption zone
Outlet, C
eB
Time, 8
QE
F igu re 3. T ypical breakthrough carve.
For most air pollution control applications, the third and fourth are the rate limiting steps.
The shape of the breakthrough curve can be predicted by the Michaels (1952) graphical
method, or analytically based on the Thomas (1944) solution.
2 3 .1 Method of Michaels (1952)
Consider the idealized breakthrough curve shown in Figure 3 resulting from flow
o f an inert gas through an adsorbent bed with a rate G’ kg/nr-s containing an inlet solute
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
concentration of Y0 kg solute/kg inert gas. The total amount of solute-free gas that has
passed through the bed up to any time is w kg/m2 of bed cross section. The gas
breakpoint and exhaustion concentrations are denoted by YB and YE, respectively. The
total amount of solute-free gas that has passed through the bed at the breakpoint is wa, at
exhaustion it is
The adsorption zone, taken to be of constant height za, is thar part of
the bed in which the concentration profile from Yg to YE exists at any time.
If 0Oand 0E are the times required for the adsorption zone to move its own length
and down the entire bed, respectively, then,
Q _
a G
(2'9)
G'
G
( 2 - 10)
Some time is required to form the adsorption zone at the beginning of the bed
when the gas is first introduced. If dF is the time required for the adsorption zone to form,
then it follows that 0£ - 0F is the time available for the zone to move through the bed once
it is formed- In this interval, the adsorption zone travels a distance za every 0a of elapsed
time. Then, if Z is the length of the bed, and the shape o f the adsorption zone does not
change after it is established,
0„
( 2 -11)
-®F
It is now necessary to find an expression for the time required to form the
adsorption zone for use in the preceding equation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
The last part of the adsorption occurs as the zone moves out of the bed. The solute
removed from the gas in the adsorption zone is U kg/m2 of bed cross section. It is given
by
U =* J ( F 0 -Y )d w
(2.12)
wa
If all the adsorbent in the zone were saturated, the solid in the zone would contain
Y0wa kg of adsorbate/m2 . Consequently, at the breakpoint, when the zone is still inside
the column, the fractional ability of the zone to still adsorb solute is
W
j ( Y 0 - Y)dw
<h
Y0wa
i
---------------« f
Y0 wa
){
o
f
Y0 )
i
(2. 13)
wa
The formation of the adsorption zone at the beginning of the bed may be assumed
to follow the same sort of pattern as has just been described for the departure of the zone;
therefore Eq. (2.13) can be used to describe the formation of the adsorption zone.
The fraction <j> is obviously some function of the shape and slope of the
breakthrough curve, and it normally approaches a value of 50%. It is useful in
establishing a relation between the time required to establish the adsorption zone and that
required for it to advance through the bed a distance equal to its thickness. If <j>= 0, so
that the adsorbent in the zone is essentially saturated, dF at the top of the bed should be
substantially the same as 0B. Conversely, if 4>= 1.0, so that the solid in the zone contains
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
essentially no adsorbate, the zone-formadon time should be very short, essentially zero.
These limiting conditions, at least, are described by:
0 F = (l-<j»)ea
(2.14)
Substituting Eq. (2.14) into Eq.(2.11),
0
za
, _7
0£
-
1-
(
w
za--------
=7
we
(2-15)
- ( l -
If the column contains ZAcpI kg of adsorbent, where Ae is the cross-sectional area
of the bed and p, is the apparent solid density in the bed, at complete saturation the bed
would contain ZAcpJ m* kg of adsorbate, where m* is the adsorbate concentration on the
solid in equilibrium with the gaseous feed. At the breakpoint, Z - Z* of the bed is
saturated, and za of the bed is saturated to the extent of 1 —<j>. The degree of overall bed
saturation at the breakpoint, a , is thus:
ex =*
(Z - z a)ps m *A c +za ps{ l- $ )m * A c
■ —
■■
— ■
Zpsm *Ac
Z-<j>za
“■
Z
(2‘16)
In the fixed bed, the adsorption zone really moves downward through the solid, as
we have seen. Imagine, instead, that the solid moves upward through the column
countercurrent to the fluid fast enough for the adsorption zone to remain stationary within
the column as in Figure 4a. Here, the solid leaving at the top of the column shows in
equilibrium with the entering gas, and all the adsorbate is removed from the effluent gas.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
As shown in Figure 4b, the operating line for the entire adsorber passes through the
origin, and intersects the equilibrium curve at the point (F0, m*).
Over the differential height dz in the adsorption zone, the rate of adsorption is
given by
G d Y = Kra { Y - Y * ) d z
(2.17)
where Kr is the overall mass-transfer coefficient for transfer from gas to solid phase. For
the adsorption zone, therefore,
Za
Yb
f &
ff
j Y - Y * * 00
G
AT
OG
(2.18)
B
For any value of z less than za, but within the adsorption zone,
Y
f
J
Z
dY
Y -Y
*
YH
7
W ~WR
---------------- *
J
(2-19)
f-J Z —
Y-Y*
The breakthrough curve can be plotted from Eqs. (2.18) and (2.19). Correlations
are needed to estimate the overall mass-transfer coefficient in Eq. (2.18). For mass
transfer from the fluid phase to the external surface of the adsorbent particles, the
following equation can be used fro laminar flow (Venneulen, et al., 1973):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
G'
(a)
nr
*0
I
m B3
>f y B
Y=
dz
y /y / / 7 z / / t
0
m —0
(b)
8
%
X
0
Operating
line
J6P
§
■s
8
■
s
bo
r
Equilibrium
line
m
Kg adsorbate/kg adsorbent
F igu re 4. The adsorption zone.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
10.96(1- z)Pb.m Mb r Df /lel0 51 i D f i f f M
r“ ~
dp \ R T
[ ^ a j
i—
J
(2.20)
where Q is the volumetric flow rate of gas, dp is the average particle diameter, 8 is the bed
porosity, pBM is the logarithmic mean partial pressure of the inert gas (approximately
equal to the total pressure for most air pollution control applications), MB is the molecular
weight of the inert gas, and pf and Df are the fluid density and difussivity respectively.
This equation is valid for laminar flow, which occurs if
R
e
„
*10
**
(2.21)
For turbulent-flow, gas-solid contact, Treybal (1980) suggests the following correlation,
valid for Rep = 90-4,000:
12.366(1- e)PB.uMl> TDf
edpAcBT
[dpQ J
1°'°”
I I* j
(2^2)
Vermeulen, et al. (1973) recommend estimating the volumetric mass-transfer
coefficient corresponding to pore diffusion from:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
ksa
60D „o.
d}
(2.23)
P
where Dp is the diffusion coefficient inside the particle. From work on diffusion inside
catalyst particles, the diffusion coefficient in the gas-filled pores inside the particle can be
estimated from (Satterfield and Sherwood, 1963):
_L I
rJ_+jJ
Dp ’
[Dk
x
» /! ’
0
° 194*
[Z
K ” SgPp * M
(224)
where
D k = Knudsen diffusion coefficient, m2/s
r = tortuosity factor ~ 4.0
Sg = surface area per gram of solid, m2/g
pp = particle density, kg/m3
X = internal porosity
The overall mass-transfer coefficient is related to the individual coefficients in the usual
way (Treybal 1980):
_ i — _ L + - *Q. .
Kya kya m * ks a
(2 25)
2 3 .2 The Thomas solution
One of the most useful treatments of the adsorption design problem is that of
Thomas (1944). He solved the set of partial differential equations that describes the
dynamics of fixed-bed adsorption systems having Langmuir-type phase equilibrium by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
introducing a transformation of the dependent variables. Consider an operation in which a
bed of adsorbent, initially free of adsorbate, that is, m(z, 0 ) = 0 , is fed at 0 = 0 with a gas
containing Y0 kg adsorbate/kg inert gas. Eventually, the whole bed will come to
equilibrium with the gaseous feed. Then the solid will contain m* kg adsorbate/kg solid.
The gradual breakthrough of the adsorbate in the gaseous effluent, i.e., the function 7(z,
Q)/Y0, and the gradual accumulation of the adsorbate on the solid, that is, m(z, Q)/m*, are
the aims of the Thomas solution. It can be expressed as
Y
J( N I K, N T )
J( N I K , N T ) + [l - J(N, N T / K )]exp[(l - K~l ) ( N - A T)j
m
1- J [ N T , N / K )
m * = / ( N ! K , NT) + [ l - J(N, NT! ^)]exp[(l - K ' l){N - AT)]
where
2Kz
N =*7------ -------( K+DH o g
K * 1 + <2 Po
m* p s z
a
y(a,p) = l-e -+ f e ~ x r0( 2 j & ) d x
o
and I0(x) is the modified Bessel function of the first kind and order zero.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(232)
(233)
(234)
(2 3 5 )
22
Values of the J function must be estimated by numerical integration. GaussLegendre quadrature, one of the various gaussian quadrature formulas, is particularly
accurate for this purpose.
2.4 Heat and momentum transfer infixed-bed adsorbers
Heat- and momentum-transfer considerations are very important in the design and
operation of fixed-bed adsorbers. Most adsorbing devices are regenerated through
intimate contact with a hot fluid. Once the regeneration step is completed, the adsorbing
solid must be cooled to its operating temperature. This is accomplished by blowing cold
air through the bed. The heat-transfer dynamics for this process resembles the adsorption
dynamics discussed in the previous section.
The pressure drop that results when a waste gas stream flows through an
adsorption device is a very important design parameter. The energy expenditure required
to overcome the pressure drop through a poorly designed fixed-bed could account for a
significant fraction of the annual operating cost for the device.
2.4.1 Cooling of thermally regenerated fixed-bed
Consider a fixed-bed that has been thermally regenerated and is initially at a
uniform temperature 0 ^ . A cold fluid, at an initial temperature 0 ro, is forced through the
bed at a mass velocity of Gc' kg/m2-s. If the fluid inside the pores of the adsorbent is at all
times in thermal equilibrium with the solid (Ruthven, 1984), the bed temperature as a
function of time and longitudinal position, 0 ,(0 , z), is given by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
e , ° - 0 /o
( 236 )
where
^rA =
haz
i /- ’
Gc c /
_
hat
jA^Ia _ n / - ’
PsCs
„ z A.s
f=“9 ~
_
0
(237)
Aa = volumetric heat-transfer coefficient
Cy-and C, = specific heat of the fluid and adsorbent, respectively.
The volumetric heat-transfer coefficient for packed beds can be estimated by a
correlation suggested by Bradshaw (1963).
ha
_ 15(1-8) f ( l - e ) I0-5 r _2/3
Gc 'C f
dP
I
R e />
j
(
238)
where
Pr = Cf nJkf
kf = thermal conductivity of the fluid.
Equation (238) is valid for R e ^ l - s) from 20 to 10,000. The fluid properties are
evaluated at ©„ = (0 ro + ©^12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
2.4.2 Pressure drop across fixed-beds.
The pressure drop through a packed bed is a standard problem in chemical engineering,
and considerable attention has been devoted to i t It can be estimated with the following
equation (Ergun, 1952):
APs3 dpPf _ 1 5 Q ( l - e ) t L7j.
Z[L —e )(C )
(239)
The first term on the right side of Eq. (239) corresponds to a friction factor in
laminar flow; the second one corresponds to turbulent flow. Unlike in flow inside a pipe,
there is a smooth transition from the laminar to the turbulent regime for flow through a
packed bed.
A recent development in the operation of fixed-bed adsorbers is the possibility of
regenerating the adsorbent using microwave (MW) radiation instead of heating the bed
with a hot fluid such as steam.
2 3 Regeneration o f fixed-bed adsorbers using microwave radiation.
Microwaves consist of electromagnetic radiation in a particular range of
wavelengths and frequencies. The wavelengths are in the centimeter range and the
frequencies are about 3 to 300 GHz (1 GHz = 1 billion Hz); waves having such high
frequencies are also called superhigh frequencies (SHF). Microwaves have a wavelength
that is shorter than the radiation used in commercial radio broadcasting but longer than
the wavelength of infrared radiation. The boundaries between various ranges of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
frequencies are arbitrary, since properties of waves change gradually rather than abruptly
with changing frequency.
Microwaves have a number of applications, such as microwave ovens, industrial
heating, radar, and telecommunications. In the field of telecommunications, microwaves
are used to carry information for telephone and television systems. One advantage of
microwaves over ordinary radio waves is that microwaves, which have a higher
frequency, can carry more information because information capacity is proportional to
frequency A drawback of microwaves is that they pass directly through the upper
atmosphere without being reflected back to the Earth, so a signal from a transmitter
cannot normally be picked up by a receiver beyond the horizon. Transmission o f
microwaves beyond line-of-sight distances requires the construction of a network of
microwave relay stations placed about 40 km (25 mi) apart on top of tall towers situated
on hilltops or the use of communications satellites as relay stations to give an
unobstructed path between the stations. Because of their high frequency, microwaves can
be accurately directed in a narrow beam from one transmitting antenna to the receiving
antenna of the next relay station.
Microwaves are usually generated in special types of electron tubes . Like the
triode—the ordinary three-electrode tube—microwave tubes contain a cathode, anode,
and grid inside an evacuated envelope. Ordinary electron tubes can operate at frequencies
up to about 30 MHz (1 MHz = 1 million Hz). Alterations in the construction of the tube
can extend the operating frequencies into the VHF (very high frequency) range (up to
about 300 MHz), but in the UHF (ultrahigh frequency) range (300-3,000 MHz) and in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
microwave range, tubes must be designed to operate in an entirely different manner,
because the frequency is comparable to the electron transit time—the time needed for
electrons to travel between electrodes. Three important microwave tubes are the klystron,
the magnetron, and the traveling-wave tube. They solve the limitation of ordinary
electron tubes by adjusting the velocity of the electron flow through the tube.
2.5.1 Interaction of microwaves with polar/polarizable materials
For those materials that are either polar or polarizable, in the presence of a
sinusoidal e. m. f., the dipoles (permanent and induced) will attempt continuously to align
themselves with the applied field (Gibson, et al., 1988). Being analogous to an electric
current, this reorientation, known as "displacement current," is the means by which
electromagnetic energy is transmitted through the material. At low frequencies of the
applied energy there is no lag between the orientation of the dipoles and the variation of
the alternating voltage; thus, the displacement current is exactly 90° out of phase with the
e. m. f. as a result of which the Joule heating—measured by the scalar product between
current and voltage—is zero. However, if the frequency is increased to MW values, the
rotation of the dipoles will begin to lag behind the voltage oscillations resulting in the
dissipation of energy as Joule heating within the material, also known as dielectric losses.
There is no exact solution to the problem of electromagnetic heating of the
materials present in a loaded adsorbent. However, it is known that the power absorbed by
a material exposed to electromagnetic fields as a result of dielectric losses is given by the
equation (Burkholder, et al., 1986):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
P = 55.63 x 10-12 f E 2e"
(2.40)
where:
P = absorbed power, W/m 3
/ = frequency, hertz
£ = rms local field intensity, V/m
e” = relative dielectric loss factor (dimensionless)
In general, as predicted by Eq. (2.40), the heat-up rate of a material in an applied
?lectric field is proportional to the dielectric loss factor. On the other hand, the
penetration depth ( 6 ), defined as the distance from the surface of the material at which the
power decays to 1/e of its value at the surface, is inversely proportional to the dielectric
loss factor (Metaxas and Meredith, 1983):
2
(2.41)
t o "
where Xq is the wavelength of the radiation in free space, and e’ is the dielectric constant
of the material.
Experimental studies have indicated that the dielectric loss-factor of the
adsorbents dominates the heat-up rate when regenerated with MW (Price and Schmidt,
1997). However, Eq. (2.41) suggests that adsorbents with very high loss-factors, such as
activated carbon, exhibit short penetration depths, making it difficult to uniformly heat a
large bed of material. For that reason, it may be more attractive to employ adsorbents
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
with low loss-factors. For such adsorbents, the loss-factor becomes a very strong function
o f the concentration of the adsorbate on the adsorbent When the bed is saturated, it has a
high loss-factor. As the VOC is driven off a portion of the bed, the local loss-factor
decreases and little additional heating takes place in this area.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 3. Previous W ork
Process design studies have been described for a new type o f VOC recovery
system that uses microwave healing to regenerate adsorbents (Price and Schmidt, 1988a;
Price and Schmidt, 1988b). Microwave (MW) regeneration systems create a highly
concentrated effluent from which the VOCs can be recovered by condensation at near­
ambient temperatures. MWs are generated in special electron tubes, such as the klystron
and the magnetron, with built-in resonators to control the frequency or by special
oscillators or solid-state devices. MW has a wavelength range from about 1mm to 30 cm
(Metaxas and Meredith, 1983).
According to Price and Schmidt (1988a), for die microwave regeneration o f an
adsorption system with a fixed-bed operation, an adequate purge can be achieved by
either pulling a vacuum on the bed or flowing a purge gas through i t Both methods
reduce the concentration o f VOCs in the vapor-phase in the bed; for vacuum-purge
regeneration the desorption effluent is a pure solvent vapor stream at low pressure while
the stream is a mixture o f inert and VOC at ambient pressure for gas-purge regeneration.
Choosing an appropriate adsorbent is one o f the most important process decisions
for VOC recovery systems, either conventional or microwave-regenerated.
The
adsorbent must posses the selectivity to adsorb the VOCs o f interest while allowing
environmentally harmless species to pass through the bed.
Since most adsorption
applications involve humid gas streams it is necessary to either choose hydrophobic
adsorbents or make provisions for dehydrating the stream. Other factors that dictate the
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
working capacity of the adsorbents are the regenerabilty and the equilibrium sorptive
capacity of VOCs at low and high concentrations. Methods like those of Michaels (1952)
and Thomas (1944) can be used to predict the shape and location o f the breakthrough
curves and the working capacity o f the bed.
Process considerations (i.e., particle size) can also affect the working capacity o f
adsorbents.
Among the process factors that can affect the working capacity o f the
absorbents are die adsorbent/VOC heat capacity; the final regeneration temperature; the
heat desorption; process safety considerations that may restrict die use o f certain
adsorbent/VOC combinations; and dielectric considerations like how well a material
absorbs microwave energy and converts it into heat as well as how die solvent dielectric
properties influence the adsorbent
Pressure affects desorption thermodynamics and kinetics as well as size and
power consumption for MW regeneration systems. Lowering the pressure may result in
larger vacuum pumps but lower microwave power consumption and generator capacity
since reducing the pressure lowers the final temperature to which the bed must be heated
to achieve a given degree of reactivation.
After the selection o f the regeneration pressure, the adsorbent regeneration degree
depends only on the final heating temperature. A uniform gas pressure and temperature
may result in no moving heat and mass transfer zones that allow a complete desorption.
The configuration of the adsorption column for fixed-bed systems represents the center o f
the design problems since it must satisfy both sorption and electromagnetic constraints.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
Some o f the important design considerations for column selection are flow
uniformity since a non-uniform air flow can results in portions o f the bed that are
unsaturated or lead to early breakthrough; uniform electromagnetic fields as nonuniformities in electric field distribution will result in bed hot spots and consequently,
incomplete or inefficient regeneration; and column cost since the size and number o f
columns required for adsorption is dictated by the volumetric flow rate of air to be
treated.
Price and Schmidt (1988a) indicated four adsorbents commonly used adsorbing
MEK from waste air streams: Dowex Optipore, UOP hydrophoic Molsiv High Silica
Zeolite (MHSZ), Calgon BPL activated carbon and Grace Davison molecular sieve 13x.
Also, they present three MW regeneration system: (a) vacuum purge MW regeneration
with non-contamining vacuum pump with atmospheric pressure condensation; (b)
vacuum-purge MW regeneration with a condenser and auxiliary vacuum pump; and (c)
inert-purge MW regeneration with heat recovery and steam preheating.
Using the method o f Michaels (1952) to predict the shape and location o f the
breakthrough curves and the working capacity o f the bed, Price and Schmidt (1988a)
concluded that Dowex Optipore and MHSZ adsorbents are the most cost-effective
selection because the high capacity and ease o f regeneration; they are hydrophobic and
completely non-reactive. Additionally, it was found that microwave regeneration of fixed
beds favors vacuum purge rather than flowing an inert stream through the bed because of
lower microwave and refrigeration costs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In another publication, Price and Schmidt (1998b) evaluated the economic
feasibility o f fixed-bed and fluidized-bed, MW regenerated adsorption systems by
systematically comparing the capital and operating costs of the proposed systems with 10
conventional VOC control technologies- They found that the MW systems had similar
capital and operating costs to conventional steam regeneration systems and, therefore,
may present an attractive alternative for recovering water-soluble solvents. In general, the
costs o f the MW subsystem was found to be a relatively minor component o f the overall
systems costs, and the MW power requirements were within the range o f commercially
available generators, even for large emission streams.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 4. M ethodology
A process performance and economic analysis will be performed to evaluate the
effect on capital and operating costs of regeneration pressure and temperature o f a
microwave regenerated VOC adsorber. The case study chosen for analysis will be the
emissions of a large-scale industrial printing operation- The concentration and flow rate
will be representative of a plant in which the VOCs are emitted into, and collected from
the building permanent total enclosure (PTE flow conditions: 68 m3/s at 500 ppmv). The
VOC will be assumed to be methyl ethyl ketone (MEK), a common industrial solvent
w ith representative dielectric and sorption properties. The adsorbent chosen will be
Dowex Optipore, a hydrophobic, polymeric adsorbent produced by Dow Chemical. This
adsorbent exhibits much better kinetic performance (i.e., very steep isotherms and high
mass transfer rates) than other adsorbents used for MEK recovery.
Figure 5 overviews the set o f process models developed to predict the
performance o f the adsorption and regeneration cycles.
First, relationships will be
developed for the relevant dielectric, sorption, and thermodynamic properties. Detailed
knowledge of the sorption equilibrium behavior over the entire regeneration cycle is
crucial to accurately estimating the heating and purge requirements for efficient
desorption. The microwave desorption kinetics program is predicated on experimental
observations which indicate that the column follows a quasiequilibrium process.
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
Adsorption Kinetics
•
Breakthrough curves
•
LMTZ
Bed saturation %
Adsorbent Properties
• Dielectric loss fector
Heat o f adsroption
• Isotherm tenperature dependence
•
Thermodinamic properties
Regeneration Kinetics
•
Heat-up rates
* Desorption rates
System Process Model
No. and size o f adsorption beds
• Condens er and HE performance
Pressure drop
• Vacuum puny performance
Cycle tunes
• Refrigeneration system performance
MW generator capacity
• Capital and operating costs
Power delivery
F igure
S.
M odeling organization fo r m icrowave (MW) regeneration studies.
An adsorption kinetics program will also be developed following the method of
Thomas (1944) to predict the shape and location of the breakthrough curves and the
working capacity of the bed.
Finally, the system process model will incorporate the output form these programs
along with other relationships to determine the overall performance characteristics of the
system. These include sizing calculations for the adsorbent bed and other components as
well as relations for determining pressure drop, blower requirements, and power
consumption of the microwave generators, vacuum pumps, and refrigeration system. The
model also will incorporate an economic analysis to estimate the capital and operating
costs o f the individual equipment items and the total plant based on vendor quotations
and correlations in the literature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 5. Results
5 .1 B r e a k th r o u g h c a lc u la tio n s u s in g T h o m a s m e th o d
To estimate the time required for die fixed-bed column using Dowex Optipore to
achieve its breakthrough point at the given conditions, it was necessary to estimate the
Langm uir isotherm constants for the system. From equilibrium data of MEK in Dowex
Optipore at 298 K provided by Dow Chemical, it was determined that the T-angmnir
isotherm equation is
3.964p .
l+10.083/>
where kt = 3.964 kg/kg-kPa and
(5.1)
= 10.083 kPa*1.
6.0
0.0
0.0
0.5
1.0
L.5
2.0
Partial Pressure fkPa]
F igure 6. L angm uir isotherm of M EK on Dowex O ptipore (25°C).
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
A waste gas consisting of 500 ppm of MEK in air at 298 K and 101325 kPa is
considered. It flows at a rate of 67.96 m3/s trough, a bed of 03048 m depth with Dowex
Optipore as adsorbent to remove the MEK vapors. The superficial velocity is 0.4318
m/s. Table 1 presents physical and chemical properties of Dowex Optipore.
T able 1. Typical Physical and C hem ical Properties o f Dowex O ptipore.
C haracteristic
M atrix Structure
Physical Form
Particle Size
M oisture Content
BET Surface Area
Total Porosity
Average Pore Diameter
A pparent Density
Ash Content
Crush Strength
H eat Capacity
Dowex O ptipore
Macroporous Styrenic Polymer
Orange to Brown Spheres
1.5 mm
<5%
1100 m2/g
1.16 cc/g
460
0 3 4 g/cc
<0.01%
>500 g/bead
1.2552 kJ/kg-K
Using the above information, the initial and equilibrium conditions detailed in
Table 2 can be determined:
T able 2. Feed and E quilibrium C onditions Through the Beds fo r 500 ppm o f M EK
in A ir a t 298 K and 1013 kPa.
Value
P a ram e te r
157.4 m2
Bed cross-sectional area
0.0507
kPa
Initial partial pressure, pn
0512 kg/nr-s
Fluid mass velocity, G ’
1.18 x 10'5 kg/m-s
Fluid viscosity, p
561.4 kg/m3
Particle density, p
0.0012 kg MEK/kg air
Inlet M EK concentration, Yn
0.133 kg MEK/kg Optipore
Equilibrium MEK adsorptivity, m*
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
However, it is still necessary to evaluate the diffusion characteristic of the system
to estim ate the breakthrough time of the fixed-bed. The molecular diffusion coefficient D
is a property of the system dependent upon temperature, pressure, and nature o f the
components. Since experimental data on diffusivity was not available for the M EK-air
system the Wilke-Lee equation was used for estimating it (Treybal, 1980):
10^ 1.084 - 0.249,
(5.2)
where
Dm - diffusivity, m2/s
T = absolute temperature, K
Ma, M b = molecular weight of A and B, respectively, kg/kmol
p t = abs pressure, N/m2
r^g = molecular separation at collision, nm
= energy of molecular attraction = yjeAeB
k = Boltzmann’s constant
ffkTIe^g) = collision function
The values of r and e can be calculated from other properties of gases, such as
normal boiling point and molar volume. If necessary, they can be estimated for each
component empirically (Treybal, 1980)
r = 1.18uI/3
(5 3 )
(5.4)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
w h e r e v is th e m o la l v o lu m e o f liq u id a t n o r m a l b o ilin g p o in t, m V k m o l, a n d 7 * is th e
n o r m a l b o i l i n g p o i n t i n K.
Table 3 summarizes the diffusion coefficients and mass-transfer coefficients for
the MEK-air system calculated from Eqs. (2.20) to (235) assuming the typical value of
0.6 fo r the internal porosity.
T ab le 3. M ass T ransfer Coefficients fo r M EK/Air M ixture 500 ppm of M EK in A ir
a t 298 K and 101-3 kPa
Coefficient
Df
d k
Dp
k fl.
k y(l
K yd
Value
9.4504e-6 m2/s
33316e-7m 2/s
53234e-8 m2/s
500.7906 kg/m3-s
3763619 kg/m3-s
373.7348 kg/m3-s
From Eq. (2.18) and the results o f Table 3,
= 0.0014 m. A t this point, it is
possible to determine the breakthrough time using the Tomas solution, given by Eqs.
(2 3 0 ) to (235). Since this involves an iterative computational scheme, EXCEL and the
FORTRAN program included in Appendix A were used to determine the time required to
achieve the breakthrough (5.70 hr). The MW regeneration time, 0R, was fixed at 1.1 hr,
and the cooling time, 0C, at 0.13 hr following Price and Schmidt (1998a)
recommendations for MW regenerated systems.
Comparing this value with the
breakthrough dme calculated by Price and Schmidt (1998a) for the system (6.4hr), a 10
% difference between them was found. It is important to notice that Price and Schmidt
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
(1998a) used Michael’s graphical method to predict the shape and location o f the
breakthrough curve.
5-2 E ffe c t o f tem perature on adsorption equilibrium
Since adsorption is an exothermic process, the concentration o f adsorbed gas
decreases with increased temperature at a given equilibrium pressure.
Table 4 summarizes microwave regeneration kinetics (temperature and adsorbent
coverage at a given time) for the system at an equilibrium regeneration pressure o f 3 3 3
kPa (5 torr).
T able 4. Regeneration of M EK from Dowex O ptipore a t 3 3 3 kPa.
Time
[min]
1
3
5
7
10
13
15
T em perature
[°C]
38
67.6
85.4
100.4
121.4
139.4
150.2
Coverage
[g MEK/jg Dow er O ptipore]
03504
0.2584
0.174
0.1195
0.0583
0.0169
0.003
Figure 7 shows the logarithmic relation between the vapor pressure and
equilibrium partial pressure at the corresponding saturation temperature versus the
adsorbate concentrations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
400 t—
U
f
2 - 300 ±T ♦
H
Z
0 I i i ; i-1- i- i t t { i -t-f- r H i i”f i t f i i M—ii i 1I i i i i I i
0
0.1
02
03
0.4
g MEK/g Dowex Optipore
F igure 7. Adsorption o f M E K on Dowex O ptipore, effect of tem perature.
Because regeneration with MW is a quasi-equilibrium process, desorption kinetics
depend only on the bed temperature and pressure. For a desorption pressure of 3 33 kPa,
Figure 8 was generated form the equilibrium data in Figure 7. Figure 8 results form the
fact that in vacuum—purge microwave regeneration the bed is heated volumetrically and
there is no significant resistance to heat or mass transfer. The typical rate-limiting
resistances in desorption are removed as a result of the unique nature of the microwave
heating process under vacuum conditions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
a i2 -
0.06 ao4 -
0.02
-
80
100
120
140
160
Tem perature [°C]
F igure 8. Reneneration kinetics a t equilibrium pressure o f 333kP a fo r
MEK/Dowex O ptipore.
5 3
C o o l m g o f th e th e r m a lly r e g e n e r a te d b e d
Considering that microwave was used to regenerate the Dowex Optipore fixed
bed; the adsorbent was exposed to an increase of temperature up to 120°C (Price and
Schmidt, 1998). It is initially at a uniform temperature QM(120°Q . In this case, a cold
fluid (air at 20°C) is forced through the bed, at an initial temperature
at a mass
velocity of Gc’kg/m*-s. Assuming that the fluid inside the pores of the adsorbent is at all
times in thermal equilibrium with the solid, the bed temperature as a function o f tim e and
longitudinal position, 0 /0 , z), is given by Eq. (236).
For the system studied, it was needed to determine the air flow required to cool
down the bed to a temperature of 25°C. A t average temperature of the fluid,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
equal to
42
343 K, Cf is 1.0145 kJ/kg-K (Smith and Van Ness, 1987) the Prandlt num ber is 0.7 (Perry
and Chilton, 1973), and the gas density pf is 1.03 kg/m3.
Using EXCEL and the FORTRAN program of Appendix A by a triai-and-error
procedure, it was found that 7.19 m3/s of air are required to cool down the bed to 25°C in
0.13 hr.
5 .4 P r e s s u r e d r o p a c r o s s th e b e d
The pressure drop through the packed bed can be estim ated by Eq (2 3 9 ). Using
this equation, the pressure drop for the adsorption process and cooling process are 1.72
kPa and 1.10 kPa, respectively.
53
D e s ig n c o n s id e r a tio n s f o r f i x e d b e d s
The primary sizing parameter for adsorbers is the total adsorbent requirement that
directly d e te rm in e s the equipment cost, and indirectly determines the size and num ber of
adsorbers vessels, and the auxiliaries such as the system fan. This param eter incorporates
several system variables, such as ( 1) adsorption, regeneration and cooling times, (2 )
waste gas volumetric flow rate, (3) allowable pressure drop, and (4) working capacity of
each bed. The amount o f adsorbent in one bed, Wd, is given by
Wa. = Y(>G A &
mw
where mwis the working capacity of the bed, given by mw = m*cu The total carbon
requirem ent for the system, Wc. is given by:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(5 3 )
where Na, Nd are number of beds adsorbing and desorbing, respectively, at any time. The
following expression relates Na and Nd to the adsorption time, 83 , the regeneration time,
0 R, and the cooling time, 0C:
N
(5.7)
0 * + 9c
Due to transportation constraints, the maximum vessel diameter that can be shopfabricated is about 4.2 m (Vatavuk, 1990). This corresponds to a cross-sectional area of
13.9 m 2/vessel. Therefore, to handle the total flow of 67.96 m3/s at a superficial gas
velocity o f 0.43 m/s, a total o f 11 vessels are needed (Na = 11). from Eq.(5.7), Nd = 2. A
total o f 13 beds are required, each with a diameter of 4.2 m, a depth of 0 3 0 5 m , a volume
o f 4.234m3, and holding of 1,440 kg o f adsorbent per bed. The total adsorbent inventory
is Wc = 18,710 kg.
5 .6 C ostfo r the D owex O ptipore adsorption system
The cost o f modular or custom built adsorption units depends on the size of the
adsorber as measured by the total adsorbent requirement,
In this case, fo r stainless
steel custom built Dowex Optipore adsorbers, the equipment cost can be calculated from
the following equation, which is a modification of the one suggested by Benitez (1993):
EC -1 3 6 .2 6 k*6 + 45WC ,
6,400 s W c s 100,000kg
(5.8)
For this case, the estimated equipment cost is $1,485,048. This cost includes adsorber
vessels, Dowex Optipore, condenser, decanter, system fan and motor, bed cooling fan
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
and motor, instruments and controls.
The cost of microwave (MW) system is not
included.
Assuming no site preparation, buildings, land or working capital are required, an
estimation of the total capital investment required was made. Table 5 presents average
installed cost factor for custom-built adsorbers.
T able S. Average C ost Factors fo r A dsorbers.
C ost Item
Direct Cost
1) Purchased equipment cost
Adsorber, auxiliary equipment, instrumentation and
controls
Taxes and freight
Total purchased equipment cost
2) Direct installation costs
Foundations and supports
Erection and handling
Electrical
Piping, Insulation and Painting
Site preparation (SP) and building (Bldg)
Total direct cost
Indirect Installation Cost
Engineering and supervisors
Construction, field and fee
Start up and performance test
Contingencies
Total indirect cost
Nondepreciable investment
Land
Working capital
Total Capital Investment (TCI)
From Neveril, et al. (1978).
C ost F acto r
LA
0.08A
PEC = 1.08 A
0.08 PEC
0.20 PEC
0.08 PEC
0.08 PEC
As required
0.44 PEC-t-SP+Bldg
0.10 PEC
0.15 PEC
0.03 PEC
0.03 PEC
031 PEC
As required
As required
1.75PEC
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
The case studied has an equipment cost equal to $1,485,048 resulting in a total
purchased equipment cost of $1,603,852 w ithout the microwave system cost. According
to Price and Schmidt (1998b), the microwave system cost is $103,000.00 fo r MW
applicator modifications (MWam); $232,000.00 for the recovery system (vacuum pump)
(RS); and $403,000.00 for microwave system (MWS). Therefore, the TCI for the whole
system was estimated in $ 4,098,241.
Table 6 presents suggested (Vatavuk, 1990) factors for estimating fixed bed
adsorber annual cost. The adsorbent has a useful lifetime of 10 years, the same as the rest
o f the equipment, therefore replacement parts are not be needed.
According to the values above, the use o f this system results in a yearly profit o f
$262,732 which means a profit of $0.039/lb VOC recovered. Table 7 sum m arizes, in
terms o f total annual cost per unit mass of VOC removed, other control technologies for
the treatment o f MEK at 500 ppm in 67.96 m3/s studied by Price and Schmidt (1998b).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
T ab le 6. Factors and Values fo r E stim ating A dsorbers A nnual Costs,
Item
D irect Operating Cost
Operating labor
Supervisor labor
M aintenance labor
M aintenance materials
Replacement parts
Utilities
Electricity
Steam
Cooling water
W aste water treatment
Indirect Operating Cost
Overhead
Property tax
Administrative charges
Insurance
Interest rate (i)
Polymeric adsorbent
lifetim e
Capital recovery factor
(CRJF)
Capital recovery
cost(CRC)
Recovery credits
Recovered adsorbate
Solvent recovery cost
Suggested F actor
Value
0_5h/shift
15%of Operating labor
0_5h/shift
100% o f maintenance labor
*****
$12.96/h
3764
*****
0.75 gal/lb VOC
*****
$14.26/h
*****
$0.Q59kW-h
*****
$0 J20/ 1000gal
*****
60% o f the sum of all labor and
m aintenance materials
1% Total Capital Investment (TCI)
2% TCI
1% TCI
10%/yr
10 year
0.1627/yr
C R F xT C I
6,666,785 Ib/yr
$0.21/lb based on 75% of virgin
solvent cost
Solvent recovery credits
Source: Vatavuk (1990); and Price and Schmidt (1998b).
$l,400,000/yr
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
T able 7. C ontrol Technologies for th e T reatm ent of M EK a t 500 ppm in 67.96 m3/s
Studied by Price and Schm idt (1998b)«
Fluid Bed AdsorptionFluid Bed. MW
Oxidation
Regeneration
0.168
Total Annual Cost
(0 .010 )
T$/lb ofVOCl*
* Number in parenthesis refers to profit
Technology
Fixed Bed MW
Regeneration
0.058
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
.
Chapter 6 Discussion of Results
Based on the data supplied by the manufacturer for adsorption of MEK on Dowex
Optipore at ambient temperature, it was found that the Langmuir adsorption isotherm
model accurately fits the experimental observations for gas phase concentrations as high
as 20,000 ppmv. The best values of the corresponding parameters, found by least-squares
estimation after linearizing the model, were 3.964 kg MEK/kg Opdpore-kPa and 10.083
k P a 1.
The use of the Thomas solution for breakthrough calculations during the
adsorption and cooling phases of the cycle greatly simplified the calculations when
compared to the Michael graphic method. So far, the difficulty in using the Thomas
solution for these purposes has been the lack of tabulated values o f the special J function,
which is the core of this elegant solution to the problem of the dynamics of fixed-bed
mass and heat transfer. The computer program developed by Benitez (1993) and used in
this work uses Gauss-Legendre quadrature formulas to accurately evaluate this function
over the whole range of possible values of its arguments. Coupled to a spreadsheet
program, such as EXCEL, it is a powerful and easily used tool for breakthrough
calculations. Besides, it was found that—for this particular case study—the Michael
method overestimated the adsorption breakthrough time by about 10% when compared to
the e s tim a te made by the Thomas solution.
Our results demonstrate that microwave heating may offer an attractive alternative
to conventional regeneration for the types of emission stream conditions assumed in this
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
analysis. In general, the circumstances that favor the use of microwave regeneration
include the following:
•
Application. First, the solvent must be expensive enough and used in sufficient
quantity to warrant recovery, since conventional regeneration technologies generally
can provide sufficient concentration ratios to economically incinerate a flow stream.
However, even for only moderately expensive solvents, the recovery credit can offer
a large potential income and may even yield a net cost savings.
•
Solvent properties. Microwave regeneration appears to be m ost attractive for watersoluble solvents, since recovery of these compounds is problem atic for conventional
steam or inert stripping. MEK, for example, is soluble in w ater up to about 28% by
w eight Is is difficult to remove by distillation since it forms an azeotrope with water
at 12%. Tables 8 classifies some o f the most common industrial solvents according
to their suitability for microwave regeneration. In addition to water solubility,
microwave regeneration lends itself to polar solvents, such as alcohols, ketones, and
acetates, which couple well wiih the applied electric field. T he dipole moment is
given in Table 8 as a quantitative measure of polarity of the solvent molecules.
•
Emission stream concentration. Finally, microwave regeneration is more suitable for
low-concentrations (less than 1,000 ppm) emission streams. These conditions require
the high concentration ratios for solvent recovery that only can be achieved with
microwave regeneration or some other form of non-convective regeneration.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
Thus, microwave regeneration appears to be an attractive VOC control technology
under certain conditions. The microwave equipment costs are a relatively small portion o f
the total system costs and are more than offset by a decrease in the size of the recovery
system and adsorbent inventory. The total system costs appear to be roughly the same as
conventional steam tegeneration systems, but allow for the recovery of water-soluble
solvents.
T able 8. Com m on In d u strial Solvents which F avo r M W R egeneration (Price and
Schm idt, 1998b).
Solvent
Solvent
Usage Rank
MEK
Methanol
M ethlene chloride
Acetone
n-Butanol
Ethyl acetate
M IK
Isobutyl alcohol
Cyclohexanone
Chloroform
Isopropyl alcohol
n-Propyl alcohol
n-Propyl acetate
M ethyl acetate
Ethylene glycol
5
6
7
7
8
11
12
15
19
22
—
—
—
—
—
W ater
Solubility
(wt %)
28
Total
1.85
Total
73
7.7
1.7
8.7
23
0.82
Total
Total
1.6
24.5
Total
Dipole
Moment
(Debye)
33
1.7
1.8
2.9
1.7
1.7
2.8
1.7
3.2
1.1
1.7
1.7
1.9
1.7
2.2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Solvent
Cost
($/lbm)
028
030
0.62
0.68
1.40
1.42
1.05
1.46
2.40
0.68
039
122
—
—
1.85
Chapter 7. Conclusions
Based on the results of this work, we conclude the following:
•
Adsorption of MEK on Dowex Optipore at ambient temperature is acurately
described by the Langmuir adsorption isotherm model for gas phase concentrations as
high as 20,000 ppmv. The best values of the corresponding parameters, found by
least-squares estimation after linearizing the model, were 3.964 kg MEK/kg OptiporekPa and 10.083 kP a1.
•
The use o f the Thomas solution for breakthrough calculations during the adsorption
and cooling phases of the cycle greatly simplified the calculations when compared to
the Michael graphic method.
•
Microwave heating may offer an attractive alternative to conventional regeneration of
VOC adsorbers when the adsorbate is water soluble, polar, and expensive enough to
justify its recovery.
•
Dowex Optipore is an excellent VOC adsorbent for MW regeneration because of its
high adsorption capacity, ease of regeneration, and the fact that it possess a relatively
low dielectric loss-factor which implies longer MW penetration depths and therefore
more uniform heating.
•
For the case study considered, a MW-regenerated adsorption system consisting of 13
beds, 4.2 m in diameter and 0 3 0 4 m in depth, each one loaded with 1,440 kg of
Dowex Optipore adsorbent, results in a profit due to MEK recovery of $0.039/lb of
solvent recovered.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
Benitez, J., Process Engineering and Design fo r A ir P ollution Control, Prendce-Hall,
N J., 1993, 148-183.
Bradshaw, R. D., AIChE J., 1963,9,590.
Burkholder, H. R., G. E. Fanslow, and D. D. Bluhm, “Recovery of Ethanol from a
M olecular Sieve by Using Dielectric Heating,” Ind. Eng. Chem. Fundam., 1986,
2 5 ,414-416.
Cooper, C. D. and F. C. Alley, A ir Pollution Control: A Design Approach, Prindle,
W eber & Schm idt, Boston, MA, 1986.
Ergun, S., Chem. Eng. Prog., 1952,4 8 ,89.
Gibson, C., I. Matthews, and A. Samuel, “Microwave Enhanced Diffusion in Polymeric
Materials”, J. M icrowave Power and Electrom agnetic Energy, 1988, 23(1), 1728.
Gilliland, E. R. and R. F. Baddour, Ind. Eng. Chem., 1953,4 5 ,330.
Metaxas, A. C. and R. J. M eredith, Industrial M icrowave H eating, 1983, Peter
Peregrinus, London.
Michaels, A.S. “Simplified M ethod of Interpreting K inetic Data in Fixed-Bed Ion
Exchange”, Ind. Eng. Chem. 1952,4 4 (8), 1922.
Neveril, R. B., J. U. Price, and K. L. Engdahl, JAPCA, 1978,2 8 ,1269.
Perry, R. H. and C. H. Chilton (eds.), Chemical Engineers Handbook, 5th ed., McGrawHill, New York, NY, 1973.
Price, D.W. and Schmidt, P.S., “Microwave Regeneration at Low Pressure: Experimental
Kinetics Studies,” J. M icrowave Power and Electrom agnetic Energy, 1997,32(3),
145-154.
Price, D.W. and Schmidt, P.S., “VOC Recovery through Microwave Regeneration of
Adsorbents: Process Design Studies”, J. o f the A ir & Waste M anage. A ssoc.,
1998a, 4 8 ,1135-1145.
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
Price, D.W. and Schmidt, P.S., “VOC Recovery through Microwave Regeneration of
Adsorbents: Comparative Economic Feasibility Studies”, J. o f the A ir & Waste
M anage. A ssoc., 1998b, 4 8 ,1146-1155.
Roussy, G., A. Zoulalian, M. Charreyre, and J. M. Thiebaut, “How Microwaves
Dehydrate Zeolites”, /. Phys. Chem. 1984,88(23), 5702-5708.
Ruthven, D. M., Principles o f Adsorption and Adsorption Processes, W iley, New York,
NY, 1984*
Satterfield, C. N. and T. K. Sherwood, The Role o f D iffusion in C atalysis, AddisonWesley, Reading, MA, 1963.
Smith, J. M. and H. C. Van Ness, Introduction to Chemical Engineering
Thermodynamics, 4thed., McGraw-Hill, New York, NY, 1987.
Thomas, H. C., “Heterogeneous Ion Exchange in a Flowing System”, J. Amer. Chem.
Soc., 1944,66, 1664-1666.
Treybal, R. E., M ass-Transfer Operations, 3rd ed., McGraw-Hill, New York, NY, 1980.
Vatavuk, W. M-, Estim ating costs o f A ir Pollution Control, Lewis, Chelsea, MI, 1990.
Vermeulen, T., G. Klein, and N. K. Hiester, Sec. 16 in J. H. Perry (ed.), Chemical
Engineers Handbook, McGraw-Hill, New York, NY, 1973.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix A. FORTRAN Computer Program for J Function
10
FUNCTION FUNCJ (ALPHA, BETA)
PARAMETER (X I = 0.0, N = 16)
DIMENSION X(20), W(20)
IF (ALPHA*BETA .LE. 100.0) THEN
CALL GAULEG (X I, ALPHA, X, W, N)
SUM = 0.0
DO 1 0 1 = 1, N
SUM = SUM+W(I)*EXP(-X(I))*BESSI0(2.0*SQRT(BErA*X(U))
CONTINUE
FUNCJ = 1.0 - EXP(-BETA)*SUM
ELSE
IF (ALPHA*BETA .LE. 3600.) THEN
FUNCJ = 0_5*( L-ERF((SQRT(ALPHA)-SQRT(BETA)))) +(EXP(
*
<SQRT(ALPHA)-SQRT(BETA))**2))/(3.5449*((ALPHA*BETA)
*
**(0.25)+BETA**(0.5)))
ELSE
FUNCJ=03-0^*ERF((SQRT(ALPHA)-SQRT(BErA)))
END IF
END IF
END
FUNCTION BESSI0(X)
REAL*8 Y ,P1,P2,P3,P4,P5,P6,P7,
* QI,Q2,Q3,Q4,Q5,Q6,Q7,Q8,Q9
DATA P l,P 2 J 33T’4 ,P 5 ^ 6 Jy7/1.0D03^156229D03-0899424DO,L2067492D
*0,
* 0.2659732DO,0360768D-1,0.45813D-2/
DATAQ1,Q2,Q3,Q4,Q5,Q6,Q7,Q8,Q9/039894228DO,0.1328592D-1,
* 0.225319D-2,-0.157565D-2,0.916281D-2,-0.2057706D-l,
* 0.2635537D-1,-0.1647633D-l,0392377D-2/
IF (ABS(X).LT 3 .75) THEN
Y=(X/3.75)**2
BESSI0=P1+Y*(P2+Y*(P3+Y*(P4+Y*(P54-Y*(P6+Y*P7)))))
ELSE
AX=ABS(X)
Y=3.75/AX
BESSI0=(EXP(AX)/SQRT(AX))*(Q1+Y*(Q2+Y*(Q3+Y*(Q4
*
+Y*(Q54-Y*(Q6+Y*(Q7+Y*(Q8+Y*Q9))))))))
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
ENDIF
RETURN
END
FUNCTION ERF(X)
IF (X . GT. 0) THEN
ERF= I - ERFCC(X)
ELSE
ERF = ERFCC(X) -1.
END IF
END
FUNCTION ERFCC(X)
Z=ABS(X)
T=17(I.+05*Z)
ERFCC=T*EXP(-Z*Z-1.26551223+T*(1.000QZ368+-T*(37409196+
* T*(.09678418+T*(-. 18628806+-T*(.27886807+T*(-1.13520398+
* T*(1.48851587+T*(-.82215223+T*.17087277)))))))))
RETURN
END
C
C
C
C
C
C
SUBROUTINE GAULEG(X1,X2,X,W,N)
THIS SUBROUTINE CALCULATES THE ROOTS AND WEIGHT
FACTORS,W,
FOR GAUSS-LEGENDRE QUADRATURE WITH N POINTS IN THE
INTERVAL
BETWEEN XI AND X2
© 1986 by Numerical Recipes Software. Reproduced by permission.
IMPLICIT REAL*8 (A-H,0-Z)
REALM X1,X2,X(N),W(N)
PARAMETER (EPS=3.D-14)
M =(N+l)/2
XM=0_5D0*(X2+X1)
XL=0.5D0*(X2-X1)
DO 1 2 1=1,M
Z=COS(3.141592654DO*a--25DO)/(N+_5DO»
I
CONTINUE
P1=1.D0
P2=0.D0
DO II J=1,N
P3=P2
P2=PI
P1=((2-D0*J- 1.D0)*Z*P2-(J-1 .D0)*P3)/J
II
CONTINUE
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PP=N*(Z*P 1-P2)/(Z*Z-1-DO)
Z1=Z
Z=Z1-P1/PP
IF(ABS(Z-Zl).GT.EPS)GO TO 1
X(D=XM-XL*Z
X(N+1-D=XM+XL*Z
W(D=2J)0*Xiy((lJD(>-Z*Z)*PP*PP)
W(N+1-I)=W(D
12 CONTINUE
RETURN
END
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
2 329 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа