You are on page 1of 6

Chemical Engineering and Processing 38 (1999) 219224

Pressure drop of arranged packings with vertical walls


Nikolai Kolev *, Svetoslav Nakov
Institute of Chemical Engineering, Bulgarian Academy of Sciences, Acad. G. Bonche6 Str., Bl.103, 1113 Soa, Bulgaria
Received 10 August 1998; accepted 25 January 1999
Abstract
Among the known types of packings, arranged packings with vertical walls have the lowest pressure drop for one transfer unit
at comparable free volume as well as the highest efciency at a given pressure drop. Reliable equations are derived for pressure
drop determination in industrial apparatuses lled with this kind of packing. The pressure drop is presented as a sum of two
terms-pressure drop (in the vertical channels) of the packing rows themselves and local pressure drop of the boundary sections
between the rows. The experimental data for dry packing pressure drop, available in the literature and newly obtained, are
described by new equation with mean deviation of 9.3%. Equations for determination of wetted packings pressure drop up to the
loading point and above it are obtained, too. The proposed equations can also be used to determine the loading point. 1999
Elsevier Science S.A. All rights reserved.
Keywords: Packed columns; Arranged packings; Packing pressure drop; Loading point; Experiments
1. Introduction
Packings with vertical walls /PVW/ belong to the
oldest types of packing. As early as in 1880, prior to Dr
Raschig patent application, arranged ceramic rings
were applied in sulfuric acid manufacture [1]. Because
of poor distribution of liquid phase over PVW, an
additional redistribution packing layer have to be in-
stalled under the liquid phase distributor. When replace
the random packing of this layer with arranged packing
specially developed for this purpose [2,3] it becomes
possible to increase liquid and gas supercial velocities
to attain values typical for PVW. A comparison of
PVW with the most performant packings revealed that
PVW are superior in their lower pressure drop for a
mass transfer unit as well as in their higher mass
transfer coefcient for similar pressure drop [46].
The existing equations for determination of pressure
drop were found to be inapplicable to the description of
thin wall PVW [7]. A possible reason is that they do not
consider the different effect of the free packing volume
on the pressure drop in the packing channels and on
local pressure drop at the rows interface.
The present paper aims at deriving of more precise
equations for evaluation of pressure drop of packings
with vertical walls.
2. Pressure drop of dry packings
2.1. Present state
Dry packing pressure drop is to be known because it
is used in the calculation of wetted packing pressure
drop.
Jonstone and Singh [8] suggested an equation de-
scribing their own experimental data obtained with grid
packings. It is inapplicable to other packings since the
equation includes only parameters typical for grid
packings-width, height, and number of packing boards.
Based on own experimental data for grid packing
elements with jagged bottom edge, as well as on data
for arranged 505 mm Raschig rings, Zhavoronkov
[9] has proposed the equation:
C=
0.78+4.25 d
e
(F
a
/F
b
)
Re
G
0.375
. (1)
The expression
* Corresponding author.
E-mail address: kolev@bas.bg (N. Kolev)
0255-2701/99/$ - see front matter 1999 Elsevier Science S.A. All rights reserved.
PII: S0255- 2701( 99) 00009- 4
N. Kole6, S. Nako6 / Chemical Engineering and Processing 38 (1999) 219224 220
4.25 d
e
/h
(F
a
/F
b
)
1.37
Re
G
0.375
describes the local pressure drop between the packing
rows. It is seen that when the term F
a
/F
b
is increased,
the local pressure drop is reduced. According to Ramm
and Zakgeim [10], this effect has no theoretical
grounds. Kolev [11] has shown that the pressure drop is
reduced because the grid packing elements [9] have
jagged bottom edges which has not been taken into
account in the determination of the term F
a
/F
b
. The
equation suggested by Jonstone and Singh [8] is free of
this anomaly.
Ramm and Zakgeim [10] have suggested the
equation:
C=u+

116+
1.235
m
2

2.6
m

d
e
h
, (2)
where the friction coefcient u=0.0075 for ceramic
packings. The calculations have revealed [11] that the
second term in Eq. (2) becomes negative for m\0.905.
Lewa [12,13] has suggested the equation:
DP
0
=a(G
2
/z
G
), (3)
where G is the mass supercial velocity of the gas
phase. The constant (depends on the nature of the gas
and on the type and dimensions of the packing. Eqs. (2)
and (3) do not account the effect of Reynolds number
and hence they are applicable only to very high
Reynolds number.
Two equations are reported In [11] derived on the
base of all experimental data for 29 PVW of different
shape and dimension, published until 1969:
C=0.59 (h/d
e
)
0.5
m
3.13
Re
G
0.266
, (4)
C=(37/Re
G
+0.113Re
G
0.1
) (h/d
e
)
0.51
m
3.42
. (5)
A precise evaluation of two components of the total
pressure drop-pressure drop caused by friction and
local pressure drop, was associated with computations,
being rather heavy in 1969. On the other hand, experi-
mental data were rather well described by Eqs. (4) and
(5). For this reason, a further study on the separate
constituents of the total pressure drop was considered
to be inexpedient for the moment.
Our later study with thin wall packings [7] revealed
that the behavior of these new packings cannot be
described by Eqs. (4) and (5). In order to obtain a
single equation, able to predict the pressure drop of
arranged packings with vertical walls, it was necessary
to obtain additional experimental data including values
of dimensionless parameters that have not been studied
till now.
The geometric parameters of the new packings are
listed in Table 1.
The experiments were carried out in a column with
rectangular cross section 224174 mm. Packing ele-
ments of the same size formed rows superimposed one
over the other. Pressure drop was measured with U-
type differential manometer. Low values below 10 mm
water column were registered with a differential
manometer equipped with optical system and micro-
metric screw (precision 0.1 Pa).
2.2. New equation
Dry packing pressure drop DP
o
was presented as a
sum of two termspressure drop in the straight section
(the rows) DP
of
and local pressure drops of the row
interface DP
ol
:
DP
o
=DP
of
+DP
ol
. (6)
Similarly to a straight tube, P
of
can be expressed by:
DP
of
=u
H
d
e
w
2
z
2m
2
, (7)
where the friction coefcient u depends on Reynolds
number.
The local pressure drops in the packing P
ol
can be
described by the expression:
DP
ol
=n
H
h
w
2
z
2m
2
, (8)
where the local pressure drop coefcient (is a function
of the void fraction o, H/h accounts for the number of
local pressure drops in a given height of the packing H.
Using the dimensionless pressure drop C Eq. (1)in
Eqs. (6) (8) and accounting for all considerations
stated above, it can be written:
C=k
f
h
d
e

1
m
n
+k
l
Re
m
, (9)
where k
f
, k
l
, n and m are experimentally determined
constants.
The experimental data listed in Table 1 were treated
in order to identify the corresponding values of the
constants. Eq. (9) takes the form:
C=0.06
h
d
e

1
m
3.8
+7Re
0.64
, (10)
The precision of these experimental constants at 95%
statistical reliability is given below:
0.06090.0076; 7.0090.075; 3.890.37; 0.64
90.035.
Jonstone and Singh [8] results for grid packings in a
square column were not used in the determination of
the constants in Eq. (10), since in ref. [8] there is no
information about the packing void fraction. The void
fraction in [11] was calculated using the equation:
m=d
t
/(d
s
+d
t
), (11)
N
.
K
o
l
e
6
,
S
.
N
a
k
o
6
/
C
h
e
m
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
a
n
d
P
r
o
c
e
s
s
i
n
g
3
8
(
1
9
9
9
)
2
1
9

2
2
4
2
2
1
Table 1
Pressure drop of arranged packing with vertical walls
Symbol Packing Element height h Specic surf. area a No Inside diam. d Ref. Void fraction Equiv. diam. Material
(m
3
m
3
) (m
2
m
3
) (d
e
m) (m) (m)
C Rings 5050 0.052 0.050 1 117 0.743 0.0254 Ceramic [11]
Rings with across 0.050 0.050 190 0.612 2 0.013 Ceramic [11]
Honeycomb blocks 0.061 0.020 3 151 0.764 0.02 Ceramic [11]
4 4 Honeycomb blocks 0.087 0.020 154 0.769 0.02 Ceramic [11]
5 Honeycomb blocks 0.099 0.030 105 0.775 0.03 Ceramic [11]
& Honeycomb blocks 0.235 0.020 6 187 0.936 0.02 Plastic [7]
Honeycomb blocks 7 0.117 0.020 187 0.936 0.02 Plastic [7]
Honeycomb blocks 0.058 0.020 8 187 0.936 0.02 Plastic [7]
9 Honeycomb blocks 0.023 0.027 103 0.75 0.029 Ceramic new
10 * Honeycomb blocks 0.040 0.027 97.2 0.74 0.0306 Ceramic new
Honeycomb blocks 0.060 0.027 11 96.4 0.75 0.0311 Ceramic new
12 q Honeycomb blocks 0.021 0.020 166.6 0.733 0.0176 Ceramic New
Honeycomb blocks 0.040 0.020 13 154 0.763 0.0198 Ceramic New
Honeycomb blocks 0.060 0.020 153 0.75 14 0.0196 Ceramic New
Honeycomb blocks-injection [6] 0.031 0.0188}0.0193 185.5 15 0.83 0.0179 Plastic
moduled
N. Kole6, S. Nako6 / Chemical Engineering and Processing 38 (1999) 219224 222
where d
s
is the thickness of the grid boards and d
t
is the
distance between two neighboring boards. It is evident
that the above equation is valid only when the length of
the apparatus wall perpendicular to the grid boards is
exactly divisible of the sum of d
s
and d
t
or when it is
many times higher than that sum. The power of m in Eq.
(10) is rather high, which requires a precise determina-
tion of m. Applying this equation to the data of Jon-
stone and Singh [8], mean deviation of 32% was
registered, all individual errors being negative except
that for four experimental points, i.e. in this case
Eq.(10) predicts higher values than the experimental
ones. For this reason the data of Jonstone and Singh [8]
were not used in the determination of the constants in
Eq. (10).
Fig. 1 illustrates the comparison between calculated
and experimental values. The mean error of Eq. (10) is
9.3%.
3. Pressure drop of wetted packings
Zhavoronkov et al. [14] have suggested a model
representing the packing pressure drop as pressure drop
of a tube. The wetted packing pressure drop under the
loading point is described by a model similar to that for
dry packing, just accounting for the reduced void frac-
tion due to the liquid phase. The packing is regarded as
a tube of irregular shape with a constant cross section
along its height. The following expression is derived:
DP=
DP
o
(1A)
3
, (12)
where A represents the packing void fraction occupied
by the liquid phase or more precisely its inuence on
the pressure drop. The same approach could be applied
considering that packing pressure drop is expressed by
two terms-pressure drop of the straight channels and
local pressure drops. Taking into account the effect of
the void fraction on the local pressure drop (Eq. (10)) it
can be written:
DP=
DP
of
(1A)
3
+
DP
ol
(1A)
6.8
. (13)
Previous studies [7,11,15] have shown that Eq. (12) can
be applied to regimes over the loading point if the term
A is presented as a sum of two constituents:
A=A
o
+DA, (14)
where A
o
is the value of A below the loading point, A
o
accounts for the effect of the additional void fraction,
occupied by the liquid phase retained in operation
regimes above the loading point.
The gas velocity in the zones of local pressure drops
considerably exceeds the velocity in the straight chan-
nels. Thus, the additional retention in the channels
could be neglected. Introducing DA in Eq. (13) it is
obtained:
DP=
DP
of
(1A
o
)
3
+
DP
ol
(1A
o
DA)
6.8
. (15)
A
o
can be presented as a function of the following
parameters:
A
o
=f(L, v
L
, z
L
, h, a, g). (16)
Applying the dimensional analysis to the experimen-
tal data for packings pressure drop below the loading
point and excluding the negligible terms, one can derive
the following expression for A
o
:
A
o
=1.2Re
L
0.05
Fr
L
0.35
. (17)
The precision of these experimental constants at 95%
statistical reliability is given below:
1.290.24; 0.05090.22; 0.3590.017.
The mean error of Eq. (17) is 10.5%. Similarly, for DA
it is obtained:
DA=f(w
m
, L, z
G
, z
L
, d
e
, h, g) (18)
DA=0.8
w
m
2
gd
e

0.6
Fr
L
0.5
Lz
L
w
m
z
G

0.6
. (19)
The precision of these experimental constants at 95%
statistical reliability is:
0.8090.15; 0.6090.17; 0.590.12; 0.690.25.
The mean error of Eq. (19) is 22.3%. This precision is
satisfactory, because DA is only a correction of A
o
.
Fig. 2 illustrates the comparison of experimental
pressure drop of wetted packings, both up to the load-
ing point and above it, with the values calculated
through Eqs. (17) and (19).
Fig. 1. Comparison of experimental data for dry packings pressure
drop with results calculated by Eq. (10) (see Table 1 for symbols).
N. Kole6, S. Nako6 / Chemical Engineering and Processing 38 (1999) 219224 223
Fig. 2. Comparison of experimental data for wet packings pressure
drop below and above the loading point with results calculated Eqs.
(17) and (19) (see Table 1 for symbols).
Fig. 3. K
h
value as a function of
Lz
L
w
m
z
G

.
for all packings (see Table 1 for symbols).
If the friction losses are taken into account, the
calculated K
h
value will be higher and will depend on
the ratio of friction losses and local pressure drop.
Fig. 3. represents K
h
plotted versus the group
Lz
L
w
m
z
G

.
K
h
values for all packings listed in Table 1 were calcu-
lated using Eqs. (18) (20). Its average value is 1.10.As
it should be expected, packings with long channels and
thin walls, i.e. packings with relatively lower local
pressure drop as compared to friction losses, have the
highest values of K
h
(see Fig. 3).
The experimental data t fairly well to the line
calculated with K
h
=1.1. It proves that it is not neces-
sary to account for the additional effects associated
with the local pressure drop to friction loss ratio.
Acknowledgements
The nancial support of the Bulgarian Fund for
Scientic Research is gratefully acknowledged.
Appendix A. Nomenclature
a specic surface of the packing (m
1
)
effect of the part of the column free A
o
cross section occupied by the liquid on
the pressure drop, under the loading
point
diameter of the circle inscribed in the d
packing hole (m)
packing equivalent diameter,
4m
a
(m) d
e
g gravity acceleration (m s
2
)
h height of a packing element (m)
height of the packing layer (m) H
F
a
maximum free section in the layer
(m
2
m
2
)
F
b
minimum free section in the contact eld
between two neighboring rows
(m
2
m
2
)
The proposed equations can be used to determine the
loading point. It has been shown [11] that the loading
point can be dened as a point at which the term K
h
K
h
=
1A
o
1(A
o
+DA)
, (20)
becomes noticeably different from 1. Determined on the
base of a great number of experimental loading points,
the value of K
h
appears to be 1.03 [11].
The value of K
h
depends not only on experimental
data but also on the relations used to estimate A
o
and
DA in Eq. (20). In ref. [11] Eq. (11) is used. Here the
power of (1A) is 3, while Eq. (20) is derived from
Eq. (15), where the power in the expression for the local
pressure drop is 6.8. In fact, the loading point is that at
which the pressure drop calculated with account of DA
differs considerably from the pressure drop calculated
without DA. In [11] DA is determined in a way to affect
the total pressure drop, while in Eq. (15) it affects only
the local pressure drop. It is obvious that the value of
K
h
used in the proposed equation should exceed the
value of K
h
as determined in Eq. (11).
The lower limit of this value, supposing that the
losses caused by friction are negligible, can be obtained
having in mind that according to Eq. (12). K
h
is a cube
root of the ratio of the pressure drop calculated with
account of DA and that without accounting for DA at
the point where DA begins to inuence the pressure
drop. The power in the proposed Eq. (15) is 6.8 in
contrast to 3 in ref. [11]. Then the lower limit of K
h
will be:
1.03
6.8/3
=1.0693.
N. Kole6, S. Nako6 / Chemical Engineering and Processing 38 (1999) 219224 224
K
h
dened in Eq. (20)
L liquid phase supercial velocity (m s
1
)
gas phase supercial velocity (m s
1
) w
w
m
gas velocity related to packing void frac-
tion (w/m, m s
1
)
Greek letters
increase in A
o
value over the loading DA
point
DP wetted packing pressure drop (Pa)
dry packing pressure drop (Pa) DP
o
DP
of
pressure drop of the dry packing in the
straight section (Pa)
local pressure drops in the packing (Pa) DP
ol
calculated data for the pressure drop DP
cal
(Pa)
DP
exp
experimental data for the pressure drop
(Pa)
friction coefcient l
v dynamic viscosity (Pa s
1
)
m void fraction (m
3
m
3
)
z density (kg m
3
)
local pressure drop coefcient x
calcalculated data for the friction factor Ccal
equivalent to Euler number
Cexp experimental data for the friction factor
equivalent to Euler number
Dimensionless numbers
Re
G
=
w
e
d
e
z
G
v
G
liquid phase Reynolds number
Fr
L
=
L
2
a
g
liquid phase Froude number
gas phase Reynolds number Re
G
=
w
e
d
e
z
G
v
G
Re
L
=
4Lz
L
av
L
liquid phase Reynolds number
dimensionless pressure drop C=
DP
o
d
e
2Hz
G
w
m
2
Subscripts
L liquid
gas G
References
[1] G. Lunge, Handbuch der Schwefels urefabrikation und ihrer
Nebenzweige, Fridr. Vierweg and Sohn Verlag, 1916, pp. 884
886.
[2] R. Daraktschiev, N. Kolev, G. Paskalev, Geordnet aufgebaute
Fu llkorperschichten mit guter Flu ssigkeitsquerverteilung, Ver-
vahrenstechnik 15 (8) (1981) 568572.
[3] N. Kolev, El. Iststkova, R. Darakchiev, A redistribution layer
for packed columns intended for high ow velocities. Chem.
Eng. Process. 21 (1987), 7782.
[4] N. Kolev, K. Winkler, R. Darakchiev, Z. Brosh, Development of
efcient packings for columns using the theory of mass transfer,
Khim. Prom. (Russ.) 8 (1986) 4345.
[5] N. Kolev N., R. Billet, Kr. Semkov, Uber die optimale Form
von Fullkorpern und Kolonnen. FAT Science Technology 92(7),
(1990) 291.
[6] N. Kolev, R. Billet, Kr. Semkov, J. Mackowiak, Sv. Nakov, On
the optimal form of stacks in packed columns. FAT-Science
Technology 96(7) (1994), 267270.
[7] N. Kolev, Sv. Nakov, Performance characteristics of a packing
with boundary layer turbulizers, Part I: Pressure drop and
loading point., Chem. Eng. Process 32 (1993) 389395.
[8] H. Johnstone, A. Singh, Ind. Eng. Chem. 29 (3) (1937) 286297.
[9] N.M. Zhavoronkov, Hydrodynamics and Heat Transfer in
Packed Towers, (Russ.), Sovetskaia Nauka, Moscow, 1944.
[10] V.M. Ramm, A. Y. Zakgeim, (Russ.), Tr. KChTI, 223 (1961).
[11] N. Kolev, Der Druckverlust von Absorptiossaulen mit
regelmaig angeordneten Fu llko rperen, Verfahrenstechnik 3 (4)
(1969) 163169.
[12] M. Leva, Chem. Eng. 64 (5) (1957) 259262.
[13] M. Leva, Tower Packings and Packed Tower Design, Ohio,
1951.
[14] N. M. Zhavoronkov, M. E. Aerov, N. N. Umnik, Chim. Prom.
(Russ). 10 (1948) 294.
[15] Kolev N., Wirkungsweise von Fu llkorperschu ttungen , Chem.-
Ing.-Technik,v (Russ.). 48(12) (1976), 11051112.
.

You might also like