Page 222 - Pressure Swing Adsorption

. ' : !I
PRESSURE SWING ADSORPTION I
198 DYNAMIC MODELING OF A PSA SYSTEM 199
I "---v------ .
' '
--v·--

.
JO~ ¾_, 40 r ~
~ L/v •25 s
., ~--=-~ Experimental 00
u C " Oiff'uslon modQJ
0 •
·o m
a -~ Q constant 0
L
C ... vor1oble 0
(l_ I
C C LDF model
4-
C ' Q ,.. 0 ' l
• □ '
,..
m
•
X Experim1.mtoi L
a " >
Diffusion modei Q
"' 10 constant D --- 0 • 40 r -------
~ L
a variable 0 "'
, __
"" ' ,-:--, LOF model ~
...._-=--=--=--= LfVOH•37 S
"
L/v H•37 s 'v _ i " "
0
,.
o L~-.l'. :_--~ -:::.::::itS_.L ·-_j 0 '
0 2 4 6 B JO 0 2 4 6 B JO
Adsorption pressure (otm) Adsorption pn:issurQ (otm)
!al (b)
Figure 5.8 Effect of adsorotton pressure on (a) mtrogen produci ounty and Figure 5.8 (Continued).
(b) nttrogen recovery at two different L / v H ratios m a kinetically controlled PSA
0
air separation process (modified cycle) showmg the companson among the expenmen•
ta! data, the LDF model, and the diffusion model with constant and vanable
dilfusivtty. Exoenmental conditions are given in Table 5.7 and other parameters m mance based on independently measured smgle-comoonent equilibrium and
Table 5.8. (From Rel. 26.) kinetic data.
The flexibility of a PSA s1mulat10n model for accommodating vanous
cyclic operations ts determtnect mamlv by the versatility of the fluid flow
i
model. (The flexibility of the fluid flow model used here has been discussed
Industrial PSA nitrogen umts operate at pressures between 7 and 10 atm in Section 5.1.) There 1s therefore no reason to orevent the applicat1on of the
(with b!owdown to atmosphenc pressure) and at a relatively low L/voH ratlo pore diffusion model to other cycle configurations.
( < 25 s). A fairly pure nitrogen product(> 99%) is produced. This level of
punty is not predicted by either the LDF model or by tlle constant-diffusivity
Effects of Process Variables
pore diffusion model. In fact at L/voH = 25 s the mtrogen oroduct purity
profiles from the LDF model as well as the constant-diffusivity pore diffusion Having established the validity of the concentrat,on-deoendent diffusivity
model become asymptotic at an oxygen concentration of about 8%. It is model for air separation over carbon molecular sieve, the modei may now be
therefore evident that the formation of nitrogen product containing Jess than used to investigate the effects of some important operating parameters on the
l % oxygen, which ts routmeiy observed in large-scale industrial umts operat- performance of the system. Some results are summarized in Figure 5.9. For a
ing at reiatively low. L/v rat10s, 1s correctly predicted oniy when allowance Skarstrorn cycie ooerat1on with no purge, increasing adsorot1on pressure
011
is made for the concentration dependence of the diffusivity. The concentra- improves product ourity at the expense of diminishing recoverv. It is shown
tion dependence of m1cropore diffusivity evidently has a strong effect on the that duai-endect pressure eoualizat1on improves both ounty and recovery
steady-state oerfonnance of a kinetically controlled PSA separation. The over those obtained by Skarstrom cycle ooeration without purge. Desorption
vanable-diffusivity model provides a reliable a priori estimate of such oerfor- without purge does not produce a high-ourity oroctuct when the feed pres-