API Stondard 621 /1S0 23264
‘The second step in sizing a knockout drum is fo consider the effect any liquid contained in the drum can have
con reducing the volume available for vapourliquid disengagement. This liquid may resutt from
@) condensate that separates during a vapour release, or
) liquid streams that accompany @ vapour release.
‘The volume occupied by the liquid should be based on a release that lasts 20 min to 30 min. Longer hold-up
times can be required if it takes longer to stop the flow. Any accumulation of liquid retained from a prior
release (from pressure-teliet devices or other sources) should be added to the liquid indicated in items a) and b)
abave to determine the available vapour-disengaging space. If the knockout drum is used to contain large
liquid dumps from pressure-relief devices from other sources and there is ne significant fieshing, and the liquid
is rémoved promplly, itis not necessary to consider this liquid volume when determining the volume available
for vapour disengagement.
It is'Important to realize as part of the sizing considerations thal the maximum vepour release case does not
necessarily coincide with the maximum liquid, Therefore, the knockout-drum size should be determined
through consideration of both the maximum-vapour-release case as well as the release case with the
maximum amount of liquid.
The following sample calculations have been limited to the simplest of the designs (7.3.2.1, drum
configurations (a) end (b)|. The calculations for drum configurations (d) and (@) are similar, with one-half the
fiow rate determining one-half the vessel length, The normal calculations are used for drum configuration (c)
and are not duplicated here.
‘The following conditions are assumed.
— A single contingency results in the flow of 25,2 kgis (200 000 Ibyt) of a fluid with a liquid density of
1496.6 kgim? (31 Ib?) and a vapour density of 2,9 ka/m® (0,18 Ib), both at flowing concitions
— The gauge pressure is 13,8 kPa (2 psi), and the temperature is 148 °C (300 “F).
— The viscosity of the vapour is 0,01 mPas (0,01 cP).
—— The fluid equilibrium results in 3,9 kg/s (31 000 l/h) of liquid and 21,3 ka/s (169 000 Ib/h) of vapour.
In addition, 1,89 m® (00 US gal) of storage for miscellaneous drainings from the units is desired. The
‘schematic in Figure 17 applies. The croplet size selected as allowable is 300 um (0,012 in) in diameter.
118‘API Standard 621 / 180 23251
a7
Koy
1 vapour and liquid safety relief valve releases
2. evel instrument to indicate when slop and drain volume has been consumed
‘3. minimum vapour space for dropout velocity
4 liquid hold-up from safety relief valves and other emergency releases,
5 slop and drain liquid
6 tofare
7 pumpout
Figure 17 — Flare knockout drum
“The vapour rate, Ry, in actual cubic metres per second (cubic feet per second), is determined as follows:
In St units:
243
Ry 7.34 mils
29
In USC units:
169.000
SiLcLues
- 3600x018 il ia
‘The drag coefficient, C, is determined from Figure 16 using Equation (29)
. (ota
485108 0,18(0,000880) (31-018) _5 apy
(0,01)
C-Re’
| From Figure 18, C-
19API Standard 621 / ISO 28251
The dropout velocity, ue, is caleulated as follows:
In Stunits:
eae
ng 1450 | 22%800%10 x(496.0-29) | graye
25x43
In USC units:
32,2 0,000984x(31-0,18) |”*
ng =4160| 2220.000084 (91-018) )" 5 a5 75
Of8x13
‘A horizontal vessel with an inside diameter, D;, and a cylindrical length, Z, should be assumed. This gives the
following total cross-sectional area, 4y
A= pay 40)
Liquid hold-up for a 30-min release from the single contingency, in addition to the slop and drain volume, is
desired, The volume in the heads is neglected for simplicity. The liquid hold-up required, 4, in cubic metres
(cubic feet), is calculated as folows:
‘The slop and drain volume of 1,89 m3 (600 U.S gal, 66,8 f9) occupies a bottom segment as follows:
In SI units:
1
Z sae 4) a)
In USC units:
1
Age ee ) (42)
‘A total of 3,9 kg/s (21 000 Ib/h) of condensed liquids with a density of 496,6 kgm? (31 ib/t®) accumulated for
30 min occupies a cross-sectional segment [see 7.3.2.1.2 a)] as follows:
A
In St units:
3 i
ae eigxeo-e0{) (43)
In USC units
31000 30/1
ene eee 1a)
A2° s(z)
“The cross-sectional area remaining or the vapour flow is expressed as follows:
A A(Art aa) 48)
120API Standard 521 / ISO 23281
The vertical depths of the liquid and vapour spaces are determined using standard geometry and the total
num diameter, %, is calculated using Equation (46):
nahithathy (46)
where
ha Is the depth of slops and drains;
(hu +h) Is the depth ofall iquid accumulation;
by is the remaining vertical space for the vapour flow.
‘The adequacy of the vapour space is verified by determining the liquid dropout time, #, using Equation (47)
In SI units:
_{*_\4
(lz) ae
In USC unis
wy
(Sy net } 8)
where
is the liquid dropout time, expressed in seconds;
h
ty is the vertical drop available for liquid dropout, expressed in centimetres (inches);
1, is the dropout velocity, expressed in metres per second (feet per second).
‘The velocity of W vapour passes, based on one vapour pass, is determined from Equations (49) and (80) for a
volume flow rate of 7,34 mols (260 £316)
tn St units:
734) 1
i (FI wy ) -
In USC units
2601
~Plz] ”
where
A, Is the cross-sectional area, expressed in square metres (square feel),
WN is the number of vapour passes;
1, is the vapour velocity, expressed in metres per second (feet per second).
atAPI Standard 621/180 23251
“The drum length required, Lyn 18 determined as follows:
Epa = hy -9-N 6)
Imi, Shall be less then or equal to the above assumed cylindrical drum length, 1; otherwise, the calculation
shall be repeated with a newly assumed cylindrical drum length.
‘Tables 15 and 16 summarize the preceding calculations for one pass for horizontal drums with various inside
diameters to determine the most economical drum size. Drum diameters in 15 cm (6in) increments are
assumed, in accordance with standard head sizes.
Table 16 — Optimizing the size of a horizontal knockout drum (SI urits)
Trial | D,* | 2 Cross-sectional area Vertical dopth oftiquid and | @° | 9 | Ln?
No. ‘vapour spaces
m2 on
AL | 4 | te | ay | tor Partha] te |e
1 | 244 | 579 | 467 | 03 | 245 | 190 | 30 | 140 | 104 | 244 | 145 | 39 | 56
220 | 625 | 40 | 020 | 227 | x83 | 20 | 137 | 91 | 220 | 128] 48 | 62
243 | 686 | 367 | 028 | 207 | 122 | 26 | 133 | a1 | a3 | 413 | 60 | 67
198 | 7.62 | 3.08 | 0,25 | 198 | os | 27 | 128 | 70 | 198 | 098 | 75 | 74
NOTE + The data inthis tabo ao In accordance with fe example von n text forone pass vapour fw
NOTE 2 The values inthis tole are roundd-off conversions ofthe values in Tabo 1,
J> fe the assumed drum inside ameter, expressed in metres
Lis the assumed drum eylndcal length, expressed in metres,
Je ae moriqu aropout me, expressed in secends.
|é ye the vapour veloc, expressed in malios per second,
Je tin sth required drum length, expressed innate,
It can be concluded from Table 15 that
— all of the drum sizes above fulfill the design requirements;
— the most suitabie drum size should be selected according to the design pressure, material requirements
‘and corrosion allowance 2s well as layout, transportation and ather considerations;
= the choice of two-pass flow, as shown in Figure 17, is optional.
422