Académique Documents
Professionnel Documents
Culture Documents
and Numerical
Adsorption/Desorption
Investigation
in Packed
of
Sorption
Beds
Flows
H. Mohamadinejad
Boeing Aerospace
Huntsville,
Company
AL
J. C. Knox
Marshall
National
Aeronautics
Huntsville,
James.
AL
E. Smith,
Department
Ph.D
of Chemical
University
of Alabama
Huntsville,
Engineering
in Huntsville
AL
ABSTRACT
The importance
nitrogen,
dimensional
diameter
partial
computer
cylindrical
differential
equations
column
(submitted
are solved
to Separatuin
breakthrough
information
Science
results
are presented
measurements
linear driving
and Tchnology)
Additional
effects of carrier
and desorption
cm in radius is examined
non-isothermal,
concentration.
on the interactive
Experimental
with concentration
sieve 5A of 0.127
simulations.
column
on packed
gas coadsorption
experimentally
for a 22.5-cm
taken at various
radial
and Newmans's
method.
force model
is required
comparisons
of carbon
of
suggests
for accurate
test
data
by
locations.
4.5-cm
The set of
Comparison
of test
that a two-dimensional
simulation
with
dioxide,
the
of the average
model
provided
interactions.
Introduction
In a packed
and therefore
increase
in porosity
particle
diameter
infinite porous
governing
of the confining
in the packed
medium
influence.
(1,2,3).
Between
shear
effects
permeability.
force
on overall
7, the permeability
R values
influence
is generally
along
For values
to increase
in bed packing
considered
above that of an
may determine
insignificant
the
(3), although
(4).
Wall effects on packed
bed processes
matrix
results
in small rectangular
range.
columns
Examples
(CDRA
are thermal
refs),
driven
processes,
swing adsorption
wall-cooled
catalytic
since
beds where
reactors
(5) and
in a packed
carder
thermal
model
balances
adsorption
The experimental
information
results
for carbon
on the sensitivity
diameter
was developed
(radial
and desorption
dioxide
of the adsorption
during
and water
adsorption
and
adsorption
to wall effects
breakthrough
in a
in a column
data is presented
of permeability
gas
Station
on the finite-differencing
in a flow-through
and canister,
model
was also
and is described
based
and desorption
Experimental
International
gas adsorption
dimensional
of wall effects
or channeling
for
along
wall.
A computer
simulates
column.
tube to particle
radial locations
the column
the influence
gas provides
with an intermediate
various
can be significant
and
the momentum
developed
to predict
technique.
and multicomponent
equation
the effect
The model
adsorption;
are included.
of flow
A two-
channeling
on
Apparatus
Carbon
bed. Single
numerical
Dioxide
Removal
Test System
Instrumentation
of the packed
column
includes
temperature
probesandsampling
tubesformeasurements
atsorbent
material
endpoints
andone
intermediate
point.To attainapproximately
adiabatic
bedconditions,
thecolumnwasfirstcovered
with
Mansfield
Q-fiberfeltinsulation,
thenwrapped
witha thermal
blanketof Mansfield
Min-Kmaterial.For
adsorption
runs,anadditional
jacketfabricated
of STSExternal
Tankinsulating
foamwasattached.
The
insulation,
approximately
3 inches
thickinall,wasusedwithsatisfactory
results.
Instrumentation
wasprovidedfor continuous
measurement
of packedcolumnoutletCO2 andH20
concentration.
Thelocationofsensors
andotherequipment
comprising
theadsorbing
apparatus
areshown
inFigure1.Thecolumnbypass
isused
priortotheteststarttoensure
stable
columninletconditions.
Thegaschromatograph,
a Shimadzu
GC-14AwithCR601integrator,
wasusedto sampleat three
sample
portlocations
to determine
gasconstituent
volumetric
fractionsduringtheadsorption
runs.The
probedepthof thesampling
tubeat thecolumnexit wasadjusted
to obtaina radialprofileof theexit
concentration
duringaseries
ofidentical
CO2adsorption
tests.
Column
A small packed column
2) located
sorbent.
at Marshall
A 10-inch
with approximate
packing
was used
Dynamics
Test Bed
dimensions
of 2 inches
Placing 4.75 inches of glass beads at the two ends of the column
results
with up to 20 inches of
and reduced
eliminated
thermal
end effects.
Procedure
Experiments
Nitrogen
bypass
were performed
on the insulated
as it proceeded
Figure
The sorbent
is a 5A zeolite.
was switched
before exposure
The column
wave
Results
3 shows
of gas mixed
rig at MSFC.
pressure
fixed-bed
efficiency
by passing
before
through
and Discussion
at the column
diameter
midpoint
column.
indicating
that channeling
material.
has a significant
effect on the
To providea measure
of theradialvariationin axialflow, adsorption
testsof CO2andN2were
conducted
repeatedly
withtheexitsampling
tubein differentradialpositions.
Thetestresults,
shownin
Figure4, showa clearrelationship
between
concentration
andradialposition.As observed
for H20
breakthrough
results,
these
resultsindicate
greater
permeability
nearthecolumn
wall.Sincetheheater
core
of the4BMS(FourBedMolecularSievs)sorbent
bedconsists
of channels
roughlyone-halfinchin
diameter,
thechanneling
effecton CO2removalwill likelybe significant.Theresultsof thetwodimensional
modelpresented
in partII alsoconfirmthatchanneling
hasa significant
effecton this
adsorption
process.
Mathematical
Gas/Solid
The
Company
model
discussed
here
(8). Nitrogen
was developed
adsorption
Equilibrium
using
was represented
Model
equilibrium
data
provided
by the W.R
Grace
&
expression:
qmiBiPi
(1)
qi = l+BiPi
Carbon dioxide
qmiBiPi
qi -
Two approaches
Prasusnitz,
by an iterative
for multiple
procedure
PY i
i=1
t
7rAni
(RTqmiJ
(9):
(2)
1-11 n--_
of gases adsorbed
components.
exp(
equation
ni
i=n
by the Langmuir-Freundlich
l+BiPini
were correlated
For simulations
adsorption
solution
single component
with
N 2
theory
isotherm,
(IAST)
of Myers
one gets:
1=0
(3)
and
Knowing
estimated
H20,
and solution
will be obtained
based
for submision
towards
in details
correlations
predicts
adsorption
is under predicted,
systems
was
for components
multicomponent
equilibrium
adsorption
as discussed
model.
equilibria
on the adsorption
of
OuC i
Ox 2
_x
For
adsorptivities.
example,
indicates
low sensitivity
However,
at low
work
humidity
continues
levels,
to
levels,
IAST
CO z
by Finn (11).
In the bulk stream of the gas within the bed, the material
_2C i
with differing
Mass Balance
Boundary
with N 2,
(4)
gas/solid
accurately
- Dl
for multicomponent
rc is
ni
of simulation
a comprehensive
Ot
equation
pressure
l+_Bjpjnj
J
agreement
_C i
on the Langmuir-Freundlich
qmiBiPi
qi
of these
by iteration
a value
on the 5a :
limitations
temperature,
to discuss
The favorable
1-e
Equation
balance
concentration
d _"
is
(5)
x< L
x< L
(6)
att>O,C
i =Ci, i forx=O
law
Ci
equation,
using an equation
P]
RT
from Edward
and knowing
and Richardson
equation
(12).
can be recast
o_P _ D
32 P
dP
_ o_
dF
(7)
azr
ar _Rrl-= ?eq-]
e i="l a
axial velocity
Gas Phase
The change
convection
of gas temperature
with respect
aTg
at
pgCpg
_ kf
a 2Tg
ax 2
in the bed.
Energy
Equation
to time is due to heat flux from the solid to the gas plus
by the following
UpgCpg
equation:
aTg
--_-x +
(8)
1 -e e hsas (T s _ rg ) -_-_-(rg
Boundary
- rw)
Tg=T
att >O,
%._=O
x< L
(9)
forx=O
for x= L
OX
in Eq 12.
Solid Phase
The energy
Energy
Equation
the term for heat flux from the solid phase to the gas
due to adsorption.
aT s
a2Ts
p sC ps -'_-t = k s 7
Boundary
d qi
(10)
Ts=T
att>-O,--_-=O
x<L
(11)
forx=O
forx=
Column
The wall temperature
Wall Energy
Equation
T w is given by:
(12)
Axial
conduction
magnitude
is neglected
(13)
since
dP _
dx
(13,14)
is used to estimate
from
This is analogous
Momentum
The Ergun equation
transfer
in a slab.
Equation
the pressure
drop,
It u- pCu 2
K
(14)
coefficient.
The empirical
bed of spherical
coefficients
developed
particle;
d2 3
K -
of
(15)
15o0- )2
C = 1.75(1- e)
dE 3
Using gas velocity
(16)
as calculated
Solid Phase
Mass
transfer
of solute
adsorbed
quantity
transfer
mechanism
adsorption
diffusion
of an adsorption
rates as discussed
amount of binder.
sorbents
Transport
Equation
Zeolite
diffusion.
state
includes
Adsorption
is driven
at equilibrium
four
steps:
by the difference
conditions.
fluid-film
since
in the actual
In general,
diffusion,
pore
it is much
greater
the mass
diffusion,
than the
by Yang, (15).
consist of crystals,
Macropore
(spaces
between
and micropore
(intracrystalline)
with a small
diffusion
must
ingeneral
beconsidered.
However,
formolecules
inthesizerangeofCO2adsorbing
onto5Athediffusion
rateinsidethepellethasbeenshown
tobecontrolled
byintracrystalline
diffusion(16).
Assuming
theintracrystalline
diffusiongoverns
the overallmasstransfer,
thelineardrivingforce
(LDF)modelbased
onGlueckauf,
(17)maybeused:
--_-=kefas(q;-qi)
wherekej
is obtained
assuming
a linear
established
by experimental
diving
force
by other researchers
Gas to pellet
and N2 x were
coefficients
(17)
to model
the adsorbed
heat transfer
.0875,
procedure
.1002,
coefficients
and
.0844
concentration
in the solid
respectively.
mass transfer
The
empirically
mass transfer
rate from the gas phase to the pellet surface, the latter was neglected.
Numerical
and equilibrium
heat balance
In
this
equation
work,
approximation
were
used
simultaneously
steep gradient
50 nodes
solution
point.
step.
the solution
numerically.
of several
differential
isotherms,
the overall
PDE's
were
spacial
coordinates.
by an implicit
method.
of mass concentration
represented
progressed.
The
equations:
the
The method
length.
was assumed
mass transfer
smaller
mixture,
the momentum
the numerical
equations,
equation,
n rate
and the
wall.
finite
difference
method.
The
first
order
discretized
finite
of Newman
(19,20)
the column
Convergence
using
of
equation,
set
well
For an n component
mass balance
discretized
of magnitudes
the
of
Solution
must be obtained
coupled
has been
obtained
of this system
phase
and Thodos
respectively.
The solution
to name a few.
obtained
cm/sec,
surface
difference
was found
1967and
equations
was
to be effective
solved
for the
1968.
from 3.6 to 60 seconds
as the
when (C_I+C_)/C "*l was less than 1.0xl0 -4 for each grid
n indicates
the previous
time
Results
In this section
we present
critical adsorption
on an ALPHA
sorbent
processes,
modeling
and comparisons
VAX computer.
is a 5A zeolite.
results
illustrate
of model
Experiments
Nitrogen
which
results
were performed
is the carrier
the importance
of accurately
gas. Experiments
begin
simulating
were performed
fixed-bed
rig at MSFC.
at ambient
The
room
temperature.
Thermal
Heat transfer
coefficients
was heated
canister
temperatures
Resultant
described
nitrogen
centerline
at 6.2
clearly
model
with heating
of the sub-scale
of 350F
and passed
heat transfer
coefficients
through
an initially
cold
column.
The
values
to the
below.
of not neglecting
mmHg
was analyzed
with simulations
of the thermal
Nitrogen
The importance
Modeling
5.
to a temperature
simulations
Effects
the nitrogen
was passed
through
periodically
at the column
has a noticeable
Coadsorption
effect;
Effects
co-adsorption
an initially
clean
inlet, midpoint,
however
Modeling
the simulations
column.
and outlet.
(dashed
in Figure
The
gas
These results
at the column
are compared
that included
N2 adsorption
6. CO 2 in
Nitrogen
over-predicted
the
effect slightly.
The thermal
sensitivity
heat of adsorption
of nitrogen
to N2 co-adsorption
through
is included.
A decrease
due to N 2 displacement
desorption.
is shown
in Figure
7. The agreement
in temperature
occurs
Water
Verification
illustrate
8 illustrates
All figures
midpoint
pressure
diameter
roughly
inch in diameter,
the two-dimensional
adsorption
process.
as adsorbed
Finally,
Since
shown
Figure
should
midpoint
indicating
the heater
here)
off by water,
be larger
outlet
than
material.
is significant
sorbent
bed consists
will be significant.
channeling
at the column
of the sorbent
that channeling
that
confirm
data as lines.
COz is driven
the channeling
(not
and simulation
The effect
at the column
through
column.
model
of the two-inch
one-half
co-adsorption
temporarily.
through
Effects Modeling
gas breaks
rises
9 shows
CO2]I--I20]N
pressure
Figure
above
Coadsorption
has
a significant
of channels
The results
effect
of
on this
and outlet.
Bed Regeneration
Desorption
of CO 2 test results
fiats out
reaches
coefficient
the effluent
in comparison
concentration
predicts
in comparison
of H20
and temperature
phase.
coefficient
Even
mass
as large
there
transfer
and
and breakthrough
adsorption,
of CO2/N 2
transfer
11 and
both temperature
was started
drop in temperature
process
are shown
Desorption
with model
shouldn't
coefficient
be such a large
the desorption
10
of H20
with .0035
difference
of H20
from
13
model are
partial pressure
in gas
between
to
increase
the H20 partialpressure
in the gasphase.In contrastwith adsorption
process,
anysmall
discrepancy
of H20partialpressure
withtestdatawill remainasaerrorthroughout
thecompletion
of the
test.In adsorption
anysmallerroratany pointin thebed,if it is caused
bytheisotherm
atsomepartial
pressure
ofthefeedwill beeliminated
ata latertimebecause
of thecorrectvalueof isotherm
ata larger
partialpressure
ofthefeed.Thiscanbeseen
fromtheearlybreakthrough
observed
byotherresearcher.
The
obvious
reason
isthattheequilibrium
isotherm
atlowpartialpressures
arebeingunderestimated.
It is also
possible
thattheequilibrium
isotherm
of H20 on 5A materialshowshysteresis
so thatthe apparent
equilibrium
pressure
observed
inadsorption
anddesorption
experiment
aredifferent.
Theconcentration
ofa
keycomponent,
CO2,is affected
by thepresence
of thenon-key
component,
N2,in CO2/N2
adsorption.
CO2effluentconcentration
overshoots
itsinletconcentration
because
of H20displacement
(moreeasily
adsorbed)
component.
Theheightof this roll-upis increased
with the inletconcentration
of H20
component.
Themostsignificant
contribution
tothedifference
in modelandexperimental
resultsof H20
desorption
pretend
to thatfactthattheduration
of anH20adsorption
runtakesaboutfifteenhoursfor
completion
of a test.Also,thedesorption
runduration
takestentotwelvehours.Duringthislongduration
thetemperature
of location
wherethetesttookplacevariedabout10-15degree
F,duringthenight.Thisin
turnaffectsthesaturated
airthatwasusedtosaturate
thecolumn.
Therefore
it is accurate
toconclude
the
amount
ofH20adsorbed
onthebedislessthanwhatisassumed
inmodel.
Onthebasisof thedatapresented
here,andothercomparisons
between
desorption
testdataand
simulation
results
notyetpublished,
thiscomputer
modelmeets
itsprimaryobjective
- achieving
predictive
capability.
Enhancements
tothemodelasdiscussed
shouldincrease
itsaccuracy.
Effortsarecontinuing
to
develop
ontheintegrated
4BMSsimulation,
equilibrium
isotherms,
heatandmasstransfer
coefficients,
and
verification
data.
Conclusions
Based
following
on
analytical
conclusions
The experimental
investigation
of convective
flows
in porous
media
the
are drawn:
results from the laboratory
the one-dimensional
obtained
and experimental
by test result
model
at the column
deviates
scale-fixed
center.
bed adsorber
The average
center
concentration
concentration
11
are quantitatively
consistent
of cross
appreciably.
sectional
This indicates
with
bed
the
NOMENCLATURE
Surface area ft 2
Langmuir
Constant
Cie
Cip
Volume
c;
of ith component
C_,o
of ith component
at boundary
of pellet fi2/ft3
constant
in Darcy equation
average
Ib mole/ft 3
of ith component
Heat capacity
cp,
Heat capacity
of solid particle,
Btu/lbm-R
Cpw
Heat capacity
of column
Btu/lbm-R
diffusivity,
DI
Axial diffusion,
no
Effective
Hw
wall,
of ith component
ib mole/ft 3
i, lb mole/ft 3
or initial lb mole/ft 3
ft2/hr
ft2/hr
Heat transfer
coefficient
between
for column
insulation,
Btu/ft2-hr
12
wall, Btu/ft2-hr
ns
Heat transfer
coefficient
between
AH
Heat of Adsorption,
Constant
K:
axial conductivity
Ks
Solid thermal
Mi
Molecular
Number
Total pressure,
mmHg or ibf/ft 2
Pi
Partial
of component
Amount
qi
qmi
Langmuir
Ri
Ro
Outside
wall diameter
Rp
Particle
radius,
Time,
Temperature
To
Ambient
Btu/lb mole
in Darcy equation
of fluid flow,
conductivity,
weight
Btu/ft-hr-R
Btu/ft-hr-R
of adsorbate
i, lb/Ib mole
of component
pressure
adsorbed
i, mmHg or Ibf/ft 2
of ith component
in equilibrium
constant
555 mm Hg ft3llb mol R
of column,
ft
of column,
ft
ft
hr
R
temperature,
gas temperature,
T_
Wall temperature,
T_
Solid temperature,
Interstitial
velocity,
ft/hr
Greek Letters
Ppg
External
Density
Btu/ft2-hr
13
Ps
Density
Pw
Density
of column
/F
Constant
AH
Heat of adsorption,
wall, lb mole/ft 3
or spreading
pressure
BTU/Ib
of solid
Subscripts
ith component
Effective
Outside,
Pg
Gas phase
ps
Solid phase
Surface
total
Wall
initial
Superscripts
Average value
Equilibrium
value
REFERENCES
1-
2-
Cohen,
29, 688,
Model
a Packed Column",
1983.
A. B. "Wall
Effect in Laminar
Packed Beds,"
AIChe
Chu, C.F.,
AIChEJ.,
"Flow
in Packed
Tubes
with a Small
1989.
14
Tube
to Particle
Diameter
Ratio",
4- Tobis,J.andVortmeyer,
D.,"NearWallChanneling
EffectonIsothermal
Constant-Pattern
Adsorption,
Chem. Eng. Sci., 1363-1369,
5-
Cheng,
in Cylindrical
6-
Beasley,
Packed Tubes,"
of Radial Thermal
J. Heat Mass
7-
Mohamadinejad,
8-
"Davison
Molecular
Baltimore
Maryland.
Langmuir,
27, 1659-1669,
Dispersion
Transfer,
9-
1988.
on Fully-Developed
30, 2373-2383,
Response
of a
Forced
Convection
Packed
1987.
Bed for Thermal
Energy
Storage,"
1984.
Equilibria",
of Glass,
Chemical
Mica, Platinum,"
Division,
J. Amer.
Chem.
A. L.,
and
J. M., Prausnitz,
"Thermodynamics
of
Mixed-Gas
Adsorption,"
AICHE
J.,
11,121,1965.
11- Finn, J.E."4BMS
Adsorption
Field, California,
12- Edwards,
Characterization
Project
Status
Report",
Ames Research
Center,
Moffet
1996.
"Gas Dispersion
in Packed
Beds,"
Chem.
Eng. Sci.,
23,
109,
1968.
13- Ergun, S.,
14- Beavers,
Mech.,
and Sparrow,
36, 711-714,
York,
E.M.,
Column,"
"Non-Darcy
Flow
through
48, 89-94,
Fibrous
1952.
Porous
Media,"
J. Applied
1969.
Packed
D. M., Principle
by Adsorption
of Adsorption
Process,
Butterworths,
and Adsorption
Boston,
Processes,
1987.
John Wiley
& Sons,
Inc., New
1984.
17Glueckauf.
E., "Theory
18- Petrovic,
Chem. Fundam.,
19-Newman,
Lawrence
of Chromatography-Part
"Mass Transfer
1955.
Packed
7, 2, 274, 1968.
J., "Numerical
Radiation
Solution
Laboratory,
Of
University
Coupled,
Ordinary
of California,
Differential
Berkeley,
15
August,
Equations",
1967.
UCRL-17739,
20-Newman,
J.,"Numerical
Solution
ofCoupled,
Ordinary
Differential
Equations,"
Ind
Science",
of Bimodal
23- Sargent,
Curves,"
Chemical
Engineering
Science,"
Series
Dependence
Isotherm,"
Chemical
1989.
R. W. H. and Whitford,
in Molecular
Microporous
of Carbon
Dioxide
102, American
DC.
16
of Diffusivity
on Zeolitic
1972.
in type 5A Molecular
Chemistry
Society,
Sieve,
Washington,
i_
- TH EIIbl ()C()II'I.E
- _lJl_
TIIh,/I_I)IJC"IR
- FLOW
DI_W POINT
,UI)NTROLLIII
SEN_OR
I
I
' --I I
I
C.4_m
Bptm m
_1 ---
"_
I
,,,,,
f"
I
. SATURATOR
lip _r,
M.c;I..bVmgW
,,
_"_
. FI_
. PP_ollE
WM IgTEII
VENT
_
- VAL_,Ig
r
v
.__
Figure
1: Adsorption
17
Test System
- QC STEM
G AUG E
AA "_----
I.d
20
I-
S,T
in.
S,T
S,T
A_ "_-"--
Thermocouple
Viox AA
Prbe
Figure
__
2:
Column
A_
Sensor
and Sampling
18
Tube
Location
.....................................
ti .................
' ............
a-_-!
a
r-1
..................
"
,
[]
VI
................................
.o.
................................
o-.......
7.................
El
&
',
.......
,:,
.........
:.................
z_
3
n
t,:
n
..................
..................
'..................
'_
............
[]
[]
.................
: ..................
E1
i
[]
',
t,
',
6
Time,
Figure
3:H20
midpoint
exit
mixed
inlet
O
I
exp. ppH20,
,_...............................
Breahthrough
19
hours
for Co2/H20/N2
Coadsorption
10
o
---o---
t_
q
O
L_
Time, Hours
Figure 4:CO2
Breakthrough
for Various
20
Radial Positions
250
I
..............................................
_
.o
O: ......
o
e-
._ ---':-.... i--::::
.D _
<>..-"
'/
12_
"
O"
200
C)
:
-O"
. -o
_ - 0---- -- _F--l:_:::
o-
.........................
150
Q
,,
[.,
,,
j"
:,"
_-
.....
.....
.........
inlet temperature,
midpoint
exit
0
[]
O
10o
'L
......................
,
,'" :
r"l . ."
:
simulation
50
0.5
5: Thermal
Model
21
1.5
Time, Hours
Figure
Validation
O
o
<>
,'
i/
0
r..)
ca,.
ta.,
--
---
midpoint
exit
experimental
[]
O
midpoint
exit
.........
.........
.........
I
0.5
Figure
6: Effect of Nitrogen
1
Time, hours
Coadsorption
22
ppCO2, inlet
1.5
on CO2 breakthrough
I05
--
sim. temp.,
--
midpoint
--
exit
I00
0
<>
',
95
.........
:': ......
',,"
._,
..........
90
[]
midpoint
<>
exit
_ .........
'",,
<>
" :
experimental
sim. temp.,
temperature,
inlet, no N2 adsorption
-........midpoint
.........
",i
<>
exit
E
0
[-
'"
,,
:
85
:
:
<>
4..............
<>
"',
;-'-,:-i .............
80
I
I
75
0.5
1
Time,
Figure
7:
Heat
Transfer
Effect
23
1.5
hours
of Nitrogen
Coadsorption
inlet
3.5
-_o=i....... __
..t..............................................
i...................................
_
2.5
E
E
..............
i-_-_--i ---
.........
J
0
9
1.5
simulated
midpoint
CO2 partial
pressure,
inlet
exit
1
O
experimental
CO2 partial
pressure,
inlet
0.5
]<>
<>
, i , , ,t
-0.5
2
6
Time,
Figure
8:
CO2
breakthrough
for CO2/H20/N2
24
hours
Coadsorption
10
..
cL
3
.............
:--
----
o
m
o
zx
-1
2
breakthrough
for CO2/H20/N2
25
at inlet
exp. ppH20,
midpoint
exit
mixed
inlet
Time, hours
Figure 9:H20
sim. ppH20
midpoint
exit
Coadsorption
10
120
110
o
n
o
....................................................
[]
[]
100
2
[]
90
80
O
o
0
0 :0
70
O0
Time, hours
Figure
for CO2/H20/N2
26
Coadsorption
10
exit
"1"
E
E
q
O
O
............................................
_-/i _
[]
exp.
ppCO2
midpointSim"
ppCO2
Outlet
at
at midpoint
inlet
QI.
t-i
[]
LJ
0.5
!tzb
1.5
2.5
Time, Hours
Figure
l l : CO2
Depletion
for CO2/N2
27
desorption
3.5
78/
....
76
i ....
i
............
_ ....
i
.....
i ....
i
! _.*_"
i ....
i
_ ....
!
i ....
i
'] .............
::.............
i ...........
i ............
i .............
!...........
::
::
,_
74
............
[L
72 ...........iI
70
68
....i_, ........i
....i...........
i'-'-'--"-'-'-'-'-'-
"-'-'-'-'-'-'-'-'-
i.
-1
"-'-'--
.....,,,,,,,,,,1
Time, Hours
Effects of CO2 depletion
28
for CO2/N2
desorption
qfrl
FII_
ii1_
lift
__..............................
15
iiii
!
!
::
mixed
...........
::
__,,
0
0
Time, Hours
Figure
13:H20
Depletion
29
for H20/N2
desorption
75
70
.....
,q),
mm
65
60
8.
IKI
E-
55
50
sim. temperature,
midpoint
exit
inlet
exp. temperature,
midpoint
exit
inlet
45
-1
2
Time,
Figure
14:
Heat
Transfer
Effects
Hours
of H2Odepletion
30
for H20/N2
desorption