Page 159 - Pressure Swing Adsorption

134 PRESSURE SWING ADSORPTION I, EQUILIBRIUM THEORY 135

0.5 I I I I 111 Table 4.1. Conditions and Results of PSA Experiments with Combined Feed and
Cocurrent Blowdown 27

Run il, ilu ~ (o/c)
0.4 R,:,pl Riti-:ory
c
$ l 12.77 10.43 25.4 26.5 -1.1
C
0 2 13.18 10.95 28.4 27.1 1.3
a. 3 17.41
E 0.3 14.35 29.6 31.4 - 1.8
0 0 4 22.91 19.03 34.1 34.6 -0.5
0
5 33.73 25.12 35.9 38.5 -2.6
6 33.29 21.06 37.4 39.1 - 1.7
0.2
>-
""
w
>
0 27
() apparatus. The theory for this type of cycle was discussed m Section 4.4.4.
w O.i
"' The application was to split oxygen from air with zerilite 13X. Six experi-
.j ments were conducted m which two pressure ratios were varied mdeoen-
0.0 1 oo 1 Q3 ! dently. A summary of the conditions and results 1s given in Table 4. L
The average absoiute deviation of the light product recovery between
theory and these experiments was only 1.5%, just as it was for the exoen-
ments havmg constant pressure durmg the feed steo. This close agreement
provides additional evidence that the local equilibrium theory 1s mdeed valid,
Figure 4.13 Expenmental recovery versus predicted recovery (according to Eo. 4.27)
for oxygen (and argon) separation from air usmg SA zeolite at 45° C. 20 and ts relatively insensitive to the cycle and ooeratmg cbndit1ons.
Finally, a set of experiments has been conducted in which a nnse step was
mtroduced, m order to extract the heavy comoonent as a pure product. In
from 6.5 to 840, and the average temperatures were 45 and 60° C. It should addition, the purge steo was left mcomolete, m order to achieve high
be noted that argon, which appears as about 1 % m dry air, adsorbs nearly recovery of the light component. The ·relevant theory 1s described in Section
ident1cally to oxygen, so the targeted "oure" light product 1s actually about 4.4.5. The specific application was to split dry air to get. oxygen (with residual
95% oxygen and 5% argon. Product recovery was of primary mterest, smce It argon) and nitrogen usmg zeolite 5A as the adsorbent. in those exoenments
1
1s predicted ouantitativeiy by the theory. Results of the expenments and the a single bed was used, and pressure rat10s were vaned between 6 and 20.
predictions of Eq. 4.27 are shown m Figure 4.13. Recall that there are no Generally, it was possible to reduce the level of imourities m the products to
!
adjustable parameters in the model, so the extent of agreement is not due to about 1 % (i.e., 0 2 m N , and N 2 m 0 2 ) anct to achieve corresoonding
and oredicted recoveries was 1.5%, and the product purity averaged 99.6% I oroduct recovenes between 27% and 90%. The oredicted and exoenmental
2
emomcal fitting. The average absolute deviation between the experimental
i recoveries of both oroducts agreed well, even though ,they vaned with the
(mtrogen-free oxygen and argon). These experiments provide strong evidence i applied pressure ratio, as shown m Figures 4.!4(a) and {b).
that, for this system, the equilibnum theory 1s essentially correct. ' In addition, Figure 4.15 shows a cross-plot of the experimental resuits
!
Although the results described are reassuring, they focus on the high-pres- from those exoenments, along with results obtamed bty Sircar 22 for a very
sure feed step; the pressunzation, purge, and especially blowdown are I similar cycle. The axes depicted are product purity and recovery. Sircar's
ancillary steps. So a major Question still remams as to the validity of ' ! results show a commonly observed trend: as recovel':}' mcreases, product
equilibrmm theories when ·pressure changes are vital to the cycle, rather than · purity decreases. The data from the exoenments described, however, show
oract1cally lmmatenal. This issue was examined in two separate types of that ounty can be maintamed at high levels as recovery increases, without a
experiments. The first type of experiment looked at the accelerat10n or significant increase in oower consurnot,on.
deceleration of the shock wave durmg the feed step, in conJunctton with To conclude this sect10n, 1t appears that the equilibrium theory 1s accurate
decreasmg or mcreasing pressure, resoectiveiy. 26 Comoarmg experimental and reliable for different PSA cycles, even for relatively difficult separations.
results with predictions of the theory showed nearly perfect agreement for Parenthetically, 1t should be mentioned that other expenmental evidence,
both mcreasing and ctecreasmg pressure. Second, PSA experiments involvmg · shown m Figure 4.3, indicates that the eouilibrmm theory should be valid for
simultaneous feed and cocurrent blowdown have been conducted in a two-bed a wide range of applications, although the degree of agreement depends on