J

.
ii
i
s
).
tII
|r
Qt;.
'R.
t1.
\. a
l1OU
r
il
|s 01
':W
ir:1l
1
Desiging piping for gavit fow
Gas entrained in liquid fowing by means of gravity from
a vessel can reduce the outlet pipe's capacity and cause
fow to surge cyclically. These problems Gin be avoided by
carefully designing lo:either full-liquid or two-phase fow.
P. D. Hils, Imperial Chemical hulllsiries PLC
o Entrainment curtails liquid gravity Bow from vessels
hy raising the pressure drop (abovc that for single-phasc
flow) through the outlet piping, and reducin
g
the static
head available for overcoming li lL pressure drop. A simi
lar problem can arise wher. H liquid is near its boiling
point, or contains dissolved gas, especially if the absolute
pressure at any point in the piping ralls below atmos
pheric pressure, as occurs in a syphon.
l:onsider the Glse of liquid flowing from the bottom of
an absorption column through a pipe that has been sized
for full liquid flow (Fig. I).
When the liquid level in the column is low enough, the
liquid cntrains gas (Fig. la). The resulting increase in
pressure drop and reduction of head restrict the !1ow
rate. and the liquid level rises (Fig. Ib). Eventually, the
level rises high enough to stop entrainment (Fig. \ c).
Howcver, gas stilI in the oULlet pipe causes I hc Icvel to
ClIt.inlle to rise lllllil the gas is all swept out (Fig. Id).
Now, the outlet pipe is running fllllllow (as was assul1 led
in the design), but the static head, becming higher thn
W assumcd, creHtes excessive i 1ow. which causes the
level to (i111 until cntrainmelll ou:urs again and the cycle
is repeHted (Fig. I e).
Such oscillations can be severe. dependin
g
on s
y
slem
ICI>ll1etry. III one LlhL the peak !low from a t Hnk L7
ceeded the ca;)acity of the vacuum breaker suniciently
colla pse it.
Generl ex
p
ression for liquid fow rate
In this anide, liquid lIowrates are generll
y
ex
p
ressed
ill terms or a di mensifnless superficial volumetric flux.
}to which is defined by:
Jt = 4Q/7t2(!d)1/2 ( I)
lIere. 0
.
is the volumetric !lowrate; d is the pipe I. D.;
,lIId g is the gravitational acceleration. Eq. (\) is si milal to
IlIe Froude number. It is used in prefercnce l the
I'mude number, because the lalter's dclinilion I'arics.
depending l circulllstances. All eqllatiolls in I his articlc
JI in consistent units.
Designing for gravity fow
Three a
p
proaches t the desi
g
n of gravity tlrinDge
<\slems arc possible:
I. For full !low, with the outlet pipin
g
size based on
single-phase criteria.
2. For selfvelltillg, with the liquid velocity in the outlet
pipe kept 101\ enough to allow gas to now counter
currently to the liquid.
3. For gas entrainment, but with the system designed
to accommodate it.
In
g
enerl. the first approach can be expected to result
in the smallest pipe diameter and should be given prefer
ence. However, in many instances, it is not possible to
ensure full pipe now-in which case, the alternatives
may have 10 hc adopted.
Designing for fooded fow
To avoid gas entrainment in the full-pipe-flow design,
the liquid level in the \"esscl must alwUys be high enough
to keep I he pipe inlet floodcd. To Hchicvc this. some
form of COllt rol will he necessarv, such as via a cOlltrol
valve (Fig. 2a) l` a vert icalloop i the piping (Fig. 2b). If
the latter is used, H syphon break will be necessary
(shown in Fig. 2b), and the piping c1ownstl'eam of the
sypholl break cannot be assumed to run flLlded because
gas is likely to _~enlrained at the syphon break. or
course. either arrangement will increase the system pres
sure drop and reduce somewhat the benelits of the
lIoodecl-f1ow design.
Single-phase criteria Gll1 be applied to designin
g
sec
tions of outlet piping in which flow can be expected to be
nooded. I I' pipi ng that is certain to be Hooded is pre
ceded by H self-venting sectiln, the self-venting section's
minimulIl Icngth should be O.51+ to provide for gas dis
engagement, beforc the piping is reduced for single
phasc flow (Fig. 2c).
rite criteria for Ilooded outlets are Eg. (2) for outlets
frolll the base of vessels. and Eg. (3) for outlets from the
side of vessc:
.I < I.()(hld)

:h > 0_892(0.f/gdJo.
25
(2)
Here, I is Ihe liquid de
p
th in the vessel away frm the
t'egion of the outlet.
.Jr < (2hl)t/2:h >0.81 I (0)2Igd"
(3)
Here. It is the liquid height above the top of the outlet
away from the region of the outlet .
(:III,IH:"t. .MIN1.1NI` 5'X!V!K :, ||II
111
LKV1Y-t.lW \YlML
Oesigning unoded (self-venting) piping
Si-otlet piping-Coming off li'om the side of a vessel,
piping should b sized such that:
`..
}[<0.3:d> @3 o' (4)
d 1G
This ensures that the line will run less than half full at its
entrance. The level in the vessel away from the outlet will
be less than 0.8 d above the base of te line. The capacity
of such an overfow line can b found from Curve 1 in
Fig. 3.
0
q
0
e

t
0
Gas
l
0. L5 5 0D\|O0O in outlet,
f051|10IDg flow. LlUlO lOw D
0XC00U8 IOW out, &DO level f158S.
oo.oo.--
SS
&5
|
b. Higher liquid 0V0 t0UUC06
6DfH00OI DU flow out is still
restricted. L0v0I continues IO
|58.
Near-lwrZlmtl piping-I such a pipeline
.
will run ony
partially full, it must be inclined to provide the statr
.
head to overcome friction losses. A minimum slope qi
1:40 is recolllmended.
To avoid having the liquid carrying gas forward, ade
quate free area mllst b left in the pipe to al\
?
w as o
l
pass backward. For pipes up to 200 mm dla., hqUlU
depths should not be more than half the pipe diameter.
For larger pipes, depths up to three-fourths of the diam
eter may be possible.
&5
|
Gas

O. )QUO l0v8l DOW 5IO5 g5 flow

in. All g&9 has D00D $WeL fDD ID0
outlet !Dg, RO & syDOD $
Ofm0U. Dg0-D50 flow in out
0I Og is helped D DgD8 D0O.
Now, outlet flow exceeds inlet
flow, and level fll$ f81O.
B. LOIfDD0Ot $I&tI5gO. Liquid
flow is still DlgD helped D5yO0O.
U11Ug9 f0$9Uf0 in !Dl0\ \8
I5 I008$0O D SU@. 5I00 I$
I01UtHlHg to first stage.
C. LUO l0v0l $ now g

___
__ -""
to l8V8HI further 8\I001.
However, 5Om0 bubbles t0 still
II0O |D outlet, f0$f1CIDg 1OW.
0v0l COOIlOU0$ to |1S0. Gas O!0I
flow 5 D0gDDOg to De COOK0U.
Entrainment causes surging by increasing pressure drop in piping and lowering head in vessel Fig. 1
CMFmTLAL ENI!MEF.R1Nf SEY`EmHFR 5, I98J
I--
^"""""..

a. Control valve maintains level in vessel

I
I

I
I
I

_ --Syphon break
r====::: i .. -
-
Un flooded
..
-7 section
1/
b. Hydraulic loop BO5UfB5 minimum level
Designs ensure flooded flow in outlet piping
i
t
1
i
t
l
Fig. 2
When fow in a partially flled pipe is uniform (i.e.,
constant depth), the energy lost through friction is bal
anced by the potential-energy change clue to the inna
tion of the pipe. In sllch a case. the mean vel{cit, FL, is
related to the inclination and the depth of !lowing liquid
by Eq. (5) [1]:
VL = (32gmi)1I2Iog{[eI4.8mJ + [0.221Im(g"mi)I/]} (5)
Here, is the hydraulic mean depth (flow areawetted
perimeter); i is the inclination of the pipe from the hori
zontal; is the pipe roughness; and V is the kinematic
Vi>Lsiy.
Fig. 3 gives the volumetric capacity for established 110w
in half-full and three-quarters-full rough and smooth
pipes. The curves were calculated via Eq. (5) for pipes of
slope l:40 and a l1uid having a kinematic viscosity of
/0-0 JJ12/S (e.g., water at 20C). The absolute rou
g
hness
used for the rough pipes was 0.25 mm (moderately ruty
mild steel). The results are not very sensitive to liquid
viscosity. The capacity of a rough pipe is inL1CbCd by
about l% |`o:a totally inviscid liquid, and is only reduced
by about l0% {'[ a liquid having a kinematic viscosit} 0('
1-5 m/s. Thus, the Fig. 3 water curves can be sale!
}
'
used l'r most liquids.
Unflooded.
section \
I
\
\
I
I
I
I
r

_>0.5
C. Slfventing section precedes flooded piping
1,000
500
200
100

E
50
2-

3
.
U
20
10
5
2
r-
I
1
:f
f

..

Curve 3IA-3I4.full ,o/ugh p.pel" 1

'

`
,
"' '
I il'/ /
"-r- . ,Q
0" .

",1 / o
"

_ .. '
. D
'\'
.
'1 /q _ I ,
"` x
r
Q'
' ".
o <'
`
k
.
.
J
L
. /
|
1// / /
I
"f 'J /
I
/
/
|
I
/
/

z
I
I
II
/
I
I !
I II
| I
I
l
I
!
i
i
l
I
I
I
I
I
i
l
I ! I
"
I
50 100 150 200 250 300 350 400
Pipe dia., RH
Capacities for established
flows in unflooded pipelines Fig. 3
CHEN1IA. FNIIELR1NO SF!,m1lFR J. I'`I
II:
!E PA.
j
i
.
-
:
V

I
t
:
|
J
|

. .
j
l
.
l

|
,
:
i
4
I
I
I
I
45'
I
`
I
i
I
I'
Ii:
i
'
I ,
'1
I
:
.
t
1:
:
/\
The initial velocity in an otltlet line dcsigned t 11
half full is less than the equilibrium velocity in a pipe
having a slope of 1 .40.As the liquid accelerates down the
pipe. the liquid depth diminishes with distance t that of
the depth corresponding to the established Ilow at a
given fowrate. To mainain a constant relative depth. a
tapered pipe would be necessary. As this is impractical.
reducing the pipe diameter stepwise is recommended.
Tapered reducers should be installed to avoid sudden
disturbances in the !low.
For long lengths of pipes. the following design ap
proach is suggested:
1. Size the outlet line on the side of a vessel for Jt =
0.3 (Curve 1 of Fig. 3). If the resulting pipe size is not
standard. choose the standard size higher thall the calcu
lated size. Continue the size so choscn for at l east tell
pipe diameters.
2. Determine the pipe diameter corresponding to
half-full established flow for the required Ilowrate (using
Curve 2A or 2B of Fig. 3). Again. select the nearest
standard pipe size higher than the calculated size .
3. Reduce the pipe diameter from the outlet size to thc
established-flow size. using an eccentric reducer that will
not change the slope of the bottom of the pipe. Prefera
bly. the reducer's minimum length should be twice that
of the upstream pipe diameter.
If the foregoing procedure is followed for pipes of
I: 40 slope, the liquid depth after the reducer will not
exceed 75% of the pipe diameter.
For long. large-diameter (>200mm) inclined pipes, it
may be worth considering a second reduction down to
the size corresponding to an establishedOow relative
depth of 75%.This reduction can be made after 50 pipe
diameters (see Curve 3A or :m of Fig. 3).
For short pipe runs, the additional cost of tapered re
ducers-especially if of a gentle angle, as recolllll1emled
(which may lIot be standard) , or of lined pipemay L%
ceed the savings in going ( smaller-diameter piping. In
such cases. the entire length of the pipe should be of the
large size.
Self-venting flow in vertical pipes
Liqliid llowing vertically clown does so as an annular
!11m. In such cases. low superficial \ clocities are necessary
to avoid gas being sucked down with the liquid. Silll,;
son's suggestion or basing pipe outlet diameters !J!1 lim
iting Froude number o/" 0.3 is recommended [2J:
Jt
<
0.3
l)
Eg. (j) being the same as Eg
.
(4). pipe diameters can be
detemlined rrom Curve 1 in Fig. :1.
This approach should be adopted when gas entrain
ment is to be avoided. as when vertical pipe extends
into a vessel to below the liquid surface. or ",hen the
dowllStream piping must be designed li)r flooded !lO\.
Smaller pipe than that dictated bv Eq. (fi) can be 7
pected l cause surging.
Self-venting fow in complex systems
Little infi)rtnation is available on lInl1oo<ied (low in
svstems that include hcnds. es
p
ecially for !low chang
es
from vertical to nearlv horizontal. and \'ice V!U. Lim
ited evidence suggests t.hat e\'en if the pipe diameter is
114
chosen lur selfvcnting 1I0w (as in a prior section Ol de
signing unlloocled piping) entrainment and surgi ng
may still occur due to the effects of the bends. The de
sign recolllmendations now given are. therefore offered
only tentatively.
Bends in the horizontal (or nearly horizontal) plane
will not necessarily cause problems if the 1: 10 slope is
continued with the bend and the bend is gentle (prefera
bly, the radius equaling five diameters) .
[n the vertical plane. the number of bends should be
limited as mllch as possible. Gently sloping piping is
preferable to vertical runs. The radius of bends should
be at least live diameters.
Bends from. or t. venical sections should be sized as
fi)r vertical pipillg. Inclined piping following a vertical
section L11 he sized ()!. half-full established !low \'ia the
criteria ('r near-horizontal piping in the previous discl-
sion on designing unlloodcd (self^venting) piping.
Changes in diameter should be made by means of ;\s\'m
metric tapered reduccrs whose lengths are equal t twice
the larger diameter. and which are installed so that the
bottom of the reducer has a slope equal to that or the
piping at either end.
If entrainment is acceptable
There are lIIany occasions when it is not ncccssar\'
prevent entrainment. Sometimes. moderate surging will
not present a problem. In such cases. piping can be sized
lor smaller diameters at considerable savings.
Sometimes, sU'ging caused by gas entrainmellt L1 be
reduced hy providing a means for the gas to escape 1 a
point downstream in the out let pipe, such as via some,
type of gas"liquid se
p
arator. If this is practical. the piping
can be of smaller r!iameter. However, because it is !lot
possible t predict the extent of entrainment-and.
hence. calculate t.he pftssure drop with certaint\-'lI1'
such approach should be auopteu cautiously.
Ackowled
g
ment
The author wishes t thank Imperial Chemical lndus
tries ('or its permissioll to publish this article.
J. Matle. !tI,/iII
References
l. .\kcts. . . r;IIJI(s for 111(' h\"llriluli,: U:'5:H n s!rW draill 5 . "('\nIS illUI piPl"
|tu$. : :lajl'S(\", Sralillllcn' ()Ilkc. |'l,
". Simpoon, l.. ... SIzing jt[tHQ |ut pr{)(css plillllS. flum. I.ny.. 111111' l, 1 (lIi
The author
l. . Hills is all ('II1im'cl ill fkp:U11111'1t1
hC.'ilHI :1IIoI(,'r nml 1HH!1O :tiali1 \\"j111
IlIIpe'nal Cht'micill 1lIIllIllri('< l.| 1.1,
Hl t. , "
'
iIlJlillgIfIIl. :"lllIh\\-il"l1.
lh:htH: (:\\,H 1 1>1. EI1l;tnd\. It .. '\ iOllsly.
: |::u11nt::1.I [u1!:8 ('IIiIU'('r iII\|
\ illl 1,,1. t +il'n'/"llIlH'lJl. 111;1111
Itrlllhll
'
"hoOliIW; ;11111 gt( .s Ikil{ll. '-k
htIH a D.'IJ).) AD1.I I'h.D. 11"11111
11II1''ri:d (:ull('gl', l.iHOO,