An Easy Method To Design Gas Vapor Relief System With Rupture Disk
An Easy Method To Design Gas Vapor Relief System With Rupture Disk
An Easy Method To Design Gas Vapor Relief System With Rupture Disk
a r t i c l e i n f o a b s t r a c t
Article history: Tank discharge gas/vapor flow problems are frequently encountered in both practice and design. To
Received 2 June 2014 perform this type of design calculation, the first step is to identify whether the flow is choked or not
Received in revised form through a trial-and-error solution of an equation for adiabatic flow with friction from a reservoir through
25 February 2015
a pipe. Developing a direct method without any trial-and-error to identify a choking condition would be
Accepted 30 March 2015
helpful for expediting the flow calculations. This paper presents an easy and quick method to identify the
Available online xxx
choking of gas flow for an emergency relief system consisting of a rupture disk and vent piping. This
greatly simplifies the design calculations. The proposed method for validating the venting adequacy of
Keywords:
Gas/vapor relief
existing ERS circumvents the iteration calculation and the use of Lapple charts. Three case studies for the
Emergency relief system design of vent piping for rupture disks support the proposed method.
Choking condition © 2015 Elsevier Ltd. All rights reserved.
Rupture disk
Design calculation
http://dx.doi.org/10.1016/j.jlp.2015.03.023
0950-4230/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023
2 G. Zhao / Journal of Loss Prevention in the Process Industries xxx (2015) 1e8
increase with a further decrease in the downstream pressure through Eq. (8a) at any location in the venting line for any given
environment while upstream pressure is fixed. For a fluid flow with Mach number. Eq. (8b) is the critical mass flux under the condition
a constant pipe diameter, the choke can only occur at the pipe exit if of assuming choked flow at the exit from a pipe frictionless flow.
the Mach number at inlet of pipe is subsonic. The gas flow may be Eq. (4) can be used to establish the relationship between the
supersonic for a short distance downstream of fittings, e.g., orifices pressures at points 1 and 2:
and contractions of the cross-sectional pipe area, where the gas can
" #1
be highly accelerated. At the end of this distance the change from
P2 M1 2 þ ðk 1ÞM21
2
2. Design of a gas/vapor relief system If flow at the exit from the ERS (location 2 in Fig. 1) is choked,
then M2 ¼ 1. Eq. (11) becomes:
A schematic for a pressure relief system with a rupture disk and
kþ1
a constant pipe diameter is shown in Fig. 1. The properties of the k 1 2 k1 2 M21
1þ M1 ¼ (12)
protected tank/vessel and its contents e like the Maximum 2 k þ 1 h22
Allowable Working Pressure (MAWP) and the gas/vapor tempera-
ture e under emergency conditions are known. The designer needs The Taylor series expansion of (1þx)n is:
to calculate the downstream “state” variables e pressure, temper-
ature, density, and velocity (Mach number) e at the inlet and outlet 1
ð1 þ xÞn ¼ 1 þ nx þ nðn 1Þx2 þ / (13)
of each pipe section. 2
Generally, it is assumed that the flow initially is steady and one- Applying the first three members of this expansion to the left
dimensional, and the vent line system is assumed to be adiabatic, side of Eq. (12) gives:
with no heat losses from the gas/vapor stream. An adiabatic flow
can be either frictional or frictionless. A reversible adiabatic flow kþ1
k 1 2 k1 kþ1 2 kþ1 4
(without friction loss) is called as an isentropic flow. Also, the gas/ 1þ M1 z1 þ M1 þ M1 (14)
2 2 4
vapor is treated as a perfect gas, such that the ideal gas law applies.
The four variables of pressure, temperature, density, and velocity Thus, Eq. (12) becomes:
(Mach number) at any location in the pipe can be solved by four " 2 #
equations of continuity (mass balance), momentum balance, en- 2 1 4
ergy balance, and equation of state at any known boundary con- M41 þ 2 1 $ M21 þ ¼0 (15)
k þ 1 h2 kþ1
ditions. Such solutions (White, 2001) are shown in Table 1, where
the stagnation value (subscript o) is used as the reference condition Solving Eq. (15) to obtain the Mach number at location 1,
for isentropic flow while the critical value under sonic condition
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
(Mach number ¼ 1, denoted by asterisks, whether or not it is u u"
u 2 u #2
u 2 1 2 1 2 4
1t 1
actually reached in the pipe exit) is used as reference condition for
M1 ¼ t $ $
adiabatic flow with friction. The stagnation state has a macroscopic k þ 1 h2 k þ 1 h2 kþ1
flow velocity of zero, which can be an actual state in a reactor, or a
hypothetical state in frictional flow where a state is isentropically (16)
transformed to a stagnation state. Mass flux can be calculated However, M1 can also be solved through Eq. (7) if the total
friction resistance coefficient of both the pipe internal surface and
fittings is known, from (Cox and Weirick, 1980):
L X
N ¼ 4f þ Ki (17)
d
Fig. 2 shows the comparison of M1 from the theoretical solution
from Eq. (7) and the simplified analytical solution from Eq. (16). For
a total friction resistance larger than 0.5, which is the case for most
practical application e because the vessel exit velocity head loss is
0.5 e the relative error of comparing the theoretical solution to the
simplified solution is less than 0.5% with the k value of 1.4. The
relative error of M1 from the theoretical solution from Eq. (7) and
Fig. 1. Schematic of pressure relief system with rupture disk. the simplified analytical solution from Eq. (16) varies from 0.2% to
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023
G. Zhao / Journal of Loss Prevention in the Process Industries xxx (2015) 1e8 3
Table 1
Equations for calculating the state variables and Mach numbers along the relief line for gas/vapor relief.
Pressure k 12
Po k 1 2 k1 P 1 kþ1
¼ 1þ M (1) ¼ (4)
P 2 P * M 2 þ ðk 1ÞM 2
Density 1 1
ro k1 2 k1
r 1 2 þ ðk 1ÞM 2 2
¼ 1þ M (3) ¼ (6)
r 2 r * M kþ1
" #
Mach number relationship 1 1 kþ1 2 þ ðk 1ÞM22 M12
kN ¼ þ ln $ (7)
M12 M22 2 2 þ ðk 1ÞM12 M22
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
kþ1
kMw 2 k1
Gc ¼ Pov ðchoked frictionless flowÞ (8b)
RTov k þ 1
0.8% for the k value from 1.9 to 1.000001 for a total friction resis- Eq. (18) is established by regressing the relative error with the
tance of 0.5. A larger error would result only if N could be less than M1e. The relative error of comparing the theoretical solution to the
0.5. The largest error is 13.6% with N ¼ 0 for isentropic flow. solution from Eq. (18) is less than 2% for any resistance coefficient,
Theoretical solution of Eq. (7) is based on the perfect gas law. N, as shown in Fig. 3.
Process conditions such as vessel pressure, k, and fluid viscosity To make Equation (16) be solvable, the following inequality has
would not impact the M1 solution from Eq. (16) as well as its to be satisfied
comparison to theoretical solution by Eq. (7), only if the perfect gas
law can be applied.
However, if the solution from Eq. (16) is set as an estimation of
" 2 # 2
M1, then M1e would be given by: 2 1 4
1 $ 0 (19)
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi k þ 1 h2 kþ1
u u"
u 2 u #2
u 2 1 t 2 1 2 4
M1e ¼ t $ 1 1 $ which gives
k þ 1 h2 k þ 1 h2 kþ1
(16a)
1
and 2 2 2
h2 $ 1 þ pffiffiffiffiffiffiffiffiffiffiffiffi (20)
kþ1 kþ1
M1 ¼ M1e þ M14:8
1e (18)
If
Fig. 2. A comparison of values of M1 from the theoretical solution of Eq. (7) and the Fig. 3. A comparison of values of M1 from the theoretical solution of Eq. (7) and the
simplified analytical solution of Eq. (16) with exit choked (M2 ¼ 1) (k ¼ 1.4). simplified analytical solution of Eq. (18) with exit choked (M2 ¼ 1) (k ¼ 1.4).
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023
4 G. Zhao / Journal of Loss Prevention in the Process Industries xxx (2015) 1e8
Table 2
1 Design parameters for capacity evaluation of rupture disk.
2 2 2
h2 > · 1 þ pffiffiffiffiffiffiffiffiffiffiffiffi (21)
kþ1 kþ1 Molecule weight, kg/kmol 20
MAWP, barg 6.90
then the ERS line exit is not choked. Relieving pressure, bara 8.60
Relieving temperature, C 93.33
Total resistance, N 4.04
3. An easy way to identify a choked condition at the exit Atmospheric pressure Pa, bara 1.01
k 1.4
Based on the above discussion, an easy way to determine if a
choked condition exists at the exit from an ERS is the following
proposed procedure: With the calculated h2 from Eq. (24), the required M1 can be
calculated through Eq. (16). With known M1 and M2, the resistance
(1) Calculate h2 ¼ PPova , coefficient Nr for the required mass flux can be calculated through
1 Eq. (7). If Nr is larger than the total flow resistance of the existing
2
2 $ 1 þ pffiffiffiffiffiffiffi
(2) If h2 > kþ1 2 , ERS line exit is not choked;
kþ1 venting line, the existing ERS system is adequate. Otherwise, the
otherwise, continue with the following procedures, existing venting system is not adequate, therefore, has to be
(3) Calculate M1 from Eq. (16), modified to reduce flow resistance by reducing the length of
(4) Calculate f(M1) from the following equation (rearranged venting line, and/or enlarging the pipe/rupture disk size, and/or
from Eq. (7) with M2 ¼ 1) making flow as straight as possible, and/or reducing number of pipe
fittings, etc.
If ERS venting line exit is not choked, M1 can be calculated in the
! " # same way as above from Eq. (24) and Eq. (16) because h2 can be
1 kþ1 ðk þ 1ÞM12 calculated from Eq. (24) (however, the actual h2 ¼ PP0va ). The M2 can
fðM1 Þ ¼ 1 þ ln kN (22) also be directly solved with known P2 ¼ Pa through applying Eq.
M12 2 2 þ ðk 1ÞM12
(8a) to location 2,
if f(M1) 0, ERS line exit is choked. Otherwise, the ERS is not
sffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
choked at the exit.
kMw kMw k1 2
Two type of ERS design work are mostly required in the engi- Gr ¼ P2 M2 ¼ Pa M2 1þ M2 (25)
RT2 RTov 2
neering practice. One is for capacity evaluation. The other is for
validating adequacy of the existing ERS. Then M2 is
For capacity evaluation, the above proposed procedures (1) to
(4) can be used to identify if the ERS line exit is choked or not. If ERS vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
venting line exit is choked, h2 is not known. M1 has to be solved u 2
u1 þ 1 þ 2ðk 1Þ G2r $ RTov
from Eq. (7) with M2 ¼ 1. If ERS line exit is not choked, h2 ¼ PP0va is t P 2 kMw
known. Eq. (7) and Eq. (11) have to be solved simultaneously to get M2 ¼ (26)
k1
M1 and M2. With known Mach number of location 1 and 2 as shown
in Fig. 1, all state variables of pressure, temperature, and density at In the same way as above, with known M1 and M2, the resistance
location 1 and 2 can be calculated from Equations (1)e(6). Then the coefficient Nr for the required mass flux can be calculated through
mass flux in the ERS can be calculated from Eq. (8a). Eq. (7). If Nr is larger than the total flow resistance of the existing
For validating adequacy of the existing ERS, the required mass venting line, the existing ERS system is adequate.
flow rate, accordingly, the required mass flux, Gr, is known. The In this way, the iteration calculation and the use of Lapple charts
above proposed procedures (1) to (4) can be used to identify if the to solve Eq. (7) is circumvented, and the validation of adequacy of
ERS line exit is choked or not. If ERS venting line exit is choked with ERS is quite straightforward.
M2 ¼ 1, dividing Eq. (8a) by Eq. (8b) The following three case studies are using the proposed strategy
to evaluate the design of ERS with rupture disk.
qffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P2 kMw P2 kMw kþ1 k
Gr RT2 RTov $ 2 P k þ 1 k1 Case study 1. Capacity evaluation
¼ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 2
Gc kþ1 kþ1 Pov 2 The Annex E of API 520 (8th Edition) gives an example of ca-
k1 k1
Pov kMw 2 Pov kMw 2 pacity evaluation of rupture disk and piping system of 100% vapor
RTov kþ1 RTov kþ1
flow with constant pipe diameter. The given design parameters
(23) shows in Table 2.
Using the proposed method as above,
From Eq. (23), the required h2 ¼ h*2 ,
k 1:01
Gr 2 k1
h2 ¼ ¼ 0:118;
h2 ¼ (24) 8:60
Gc k þ 1
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u u"
u 2 u 2 # 2
u 2 1 t 2 1 4
M1 ¼ t $ 1 1 $ ¼ 0:130
1:4 þ 1 0:118 1:4 þ 1 0:118 1:4 þ 1
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023
G. Zhao / Journal of Loss Prevention in the Process Industries xxx (2015) 1e8 5
Table 3 Table 4
The calculated results for capacity evaluation of rupture disk. Design parameters for neopentyl glycol relief (Case study 2).
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u u"
u 2 u 2 # 2
u 2 1 2 1 4
M1 ¼ t $ 1t 1 $ ¼ 0:244
1:06 þ 1 0:233 1:06 þ 1 0:233 1:06 þ 1
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u u"
u 2 u 2 # 2
u 2 1 t 2 1 4
M1 ¼ t $ 1 1 $ ¼ 0:141
1:06 þ 1 0:138 1:06 þ 1 0:138 1:06 þ 1
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023
6 G. Zhao / Journal of Loss Prevention in the Process Industries xxx (2015) 1e8
Table 5 Table 6
The calculated results for neopentyl glycol relief with a 0.0762 m rupture disk The calculated results for neopentyl glycol relief with a 0.0508 m rupture disk
installed. installed.
Location 0 1 2 Location 0 1 2
Pressure P, bara 4.35 4.14 1.25 Pressure P, bara 4.35 4.16 1.18
Stagnation pressure Po, bara 4.35 4.35 2.11 Stagnation pressure Po, bara 4.35 4.35 1.98
Temperature T, C 259.35 257.86 243.84 Temperature T, C 259.35 258.05 243.84
Stagnation temperature To, C 259.35 259.35 259.35 Stagnation temperature To, C 259.35 259.35 259.35
Mach number: M 0 0.31 1 Mach number: M 0 0.29 1
Velocity u, m/s 0 64.85 209.16 Velocity u, m/s 0 60.70 209.16
Density r, kg/m3 10.23 9.76 3.03 Density r, kg/m3 10.23 9.82 2.85
Mass flux G, kg/m2s 633.25 633.25 Mass flux G, kg/m2s 596.19 596.19
Mass flow rate W, kg/s 2.60 2.60 Mass flow rate W, kg/s 1.09 1.09
required mass flux, Gr, kg/m2s 273.05 required mass flux, Gr, kg/m2s 614.37
required mass flowrate, kg/s 1.25 required mass flowrate, kg/s 1.25
Table 7
! " # Design parameters for neopentyl glycol relief (Case study 3).
1 1 kþ1 ðk þ 1ÞM12
Nr ¼ 1 þ ln Molecule weight, kg/kmol 104.15
k M1 2 2k 2 þ ðk 1ÞM12 Relieving pressure, Bara 2.27
Boling temperature at relieving pressure, C 235.6
1 1 1:06 þ 1 ð1:06 þ 1Þ0:1412 Total resistance, N 6.86
¼ 1 þ ln Vessel Volume: m3
1:06 0:1412 2 1:06 2 þ ð1:06 1Þ0:1412 45.42
fill ratio 0.85
¼ 42:6 filled volume, m3 38.61
filled height, m 4.31
wetted area, m2 54.68
The Nr is larger than 6.86 of the actual flow resistance. Therefore,
insulation factor (NFPA 30) 0.3
the existing ERS is adequate. Q, kW (NFPA 30) 647.66
The actual Mach number and state variables of location 1 and 2 Latent heat at 235.6 C, kJ/kg 559.02
can be also calculated from Equation (1) through (7). The calculated required mass flowrate, kg/s 1.16
results are shown in Table 5. The calculated mass flow rate, 2.60 kg/ Critical mass flux, Gc, kg/m2s 696.42
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u u"
u 2 u 2 # 2
u 2 1 t 2 1 4
M1 ¼ t $ 1 1 $ ¼ 0:333
1:06 þ 1 0:310 1:06 þ 1 0:310 1:06 þ 1
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023
G. Zhao / Journal of Loss Prevention in the Process Industries xxx (2015) 1e8 7
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u u"
u 2 u 2 # 2
u 2 1 2 1 4
M1 ¼ t $ 1t 1 $ ¼ 0:521
1:06 þ 1 0:447 1:06 þ 1 0:447 1:06 þ 1
1.35 kg/s, is larger than the required mass flowrate of 1.16 kg/s.
1 1:06 þ 1 ð1:06 þ 1Þ 0:5212 Therefore, that the existing ERS is adequate is also confirmed by
fðM1 Þ ¼ 1 þ ln
0:5212 2 2 þ ð1:06 1Þ 0:5212 capacity evaluation.
1:06 6:86 ¼ 5:91 < 0: Instead, if a 0.0635 m diameter rupture disk is installed (the
actual flow resistance is 7.42 in this case), that the pipe flow outlet
Then, following the strategy proposed in this paper, that the to atmosphere is not choked can be also immediately confirmed
pipe flow outlet to atmosphere is not choked can be immediately following the preceding procedure. Therefore, M2 < 1. Then
confirmed without any detailed calculations of trial and error.
Therefore, M2 < 1. Then k 1:06
Gr 2 k1 365:84=0:9 2 1:061
h2 ¼ ¼ ¼ 0:346
k 1:06 Gc k þ 1 696:79 1:06 þ 1
Gr 2 k1 254:05=0:9 2 1:061
h2 ¼ ¼ ¼ 0:240
Gc k þ 1 696:79 1:06 þ 1
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u u"
u 2 u 2 # 2
u 2 1 t 2 1 4
M1 ¼ t $ 1 1 $ ¼ 0:378
1:06 þ 1 0:346 1:06 þ 1 0:346 1:06 þ 1
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u u"
u 2 u 2 # 2
u 2 1 2 1 4
M1 ¼ t $ 1t 1 $ ¼ 0:252
1:06 þ 1 0:240 1:06 þ 1 0:240 1:06 þ 1
! " #
1 1 1 kþ1 2 þ ðk 1ÞM22 M12 Table 8
Nr ¼ þ ln $ The calculated results for neopentyl glycol relief with a 0.0762 m rupture disk
k M12 M22 2k 2 þ ðk 1ÞM12 M22 installed.
Location 0 1 2
1 1 1
¼ Pressure P, bara 2.27 2.16 1.01
1:06 0:2522 0:5432 Stagnation pressure Po, bara 2.27 2.27 1.25
Temperature T, C
1:06 þ 1 2 þ ð1:06 1Þ0:5432 0:2522 235.64 234.30 229.64
þ ln ¼ 10:16 Stagnation temperature To, C 235.64 235.64 235.64
2 1:06 2 þ ð1:06 1Þ0:2522 0:5432 Mach number: M 0 0.30 0.63
Velocity u, m/s 0 61.57 130.11
The Nr is larger than 6.86 of the actual flow resistance. Therefore, Density r, kg/m3 5.58 5.33 2.52
the existing ERS is adequate. Mass flux G, kg/m2s 328.64 328.64
The actual Mach number and state variables of location 1 and 2 Mass flow rate W, kg/s 1.35 1.35
can be also calculated from Equations (1) through (7). The calcu- required mass flux, Gr, kg/m2s 254.05
required mass flowrate, kg/s 1.16
lated results are shown in Table 8. The calculated mass flow rate,
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023
8 G. Zhao / Journal of Loss Prevention in the Process Industries xxx (2015) 1e8
Table 9 Acknowledgment
The calculated results for neopentyl glycol relief with a 0.0635 m rupture disk
installed.
The author would like to thank Richard Prugh and Keith Middle
Location 0 1 2 for reviewing this article.
Pressure P, bara 2.27 2.17 1.01
Stagnation pressure Po, bara 2.27 2.27 1.24 Nomenclature
Temperature T, C 235.64 234.38 229.97
Stagnation temperature To, C 235.64 235.64 235.64
Mach number: M 0 0.29 0.61
d Pipe internal diameter, m
Velocity u, m/s 0 59.65 126.47 f Fanning friction factor, dimensionless
Density r, kg/m3 5.58 5.35 2.52 k Specific heat ratio, dimensionless
Mass flux G, kg/m2s 319.23 319.23 K Resistance coefficient of pipe fitting, dimensionless
Mass flow rate W, kg/s 0.91 0.91
L Pipe length, m
required mass flux, Gr, kg/m2s 365.84
required mass flowrate, kg/s 1.16 M Mach number, dimensionless
Mw Molecular weight, kg/kmol
N Pipe resistance coefficient, dimensionless
P Pressure, bara
! " # T Temperature, K
1 1 1 kþ1 2 þ ðk 1ÞM22 M12
Nr ¼ þ ln $ h Pressure ratio, dimensionless
k M12 M22 2k 2 þ ðk 1ÞM12 M22 r Density, kg/m3
1 1 1 Subscripts
¼ 2
0,1,2 Location of relief line
1:06 0:378 0:7782
a Ambient conditions
1:06 þ 1 2 þ ð1:06 1Þ0:7782 0:3782
þ ln ¼ 3:66 c Critical condition
2 1:06 2 þ ð1:06 1Þ0:378 2 0:778 2
e Estimation
The Nr is less than 7.42 of the actual flow resistance. Therefore, o Stagnation condition
the existing ERS is not adequate. r Required
The actual Mach number and state variables of location 1 and 2
can be also calculated from Equations (1) through (7). The calcu- References
lated results are shown in Table 9. The calculated mass flow rate,
API standard 520, 2008. Sizing, Selection, and Installation of Pressure-relieving
0.91 kg/s, is less than the required mass flowrate of 1.16 kg/s. Devices, eighth ed.
Therefore, that the existing ERS is not adequate is also confirmed by API standard 521, 2008. Pressure-relieving and Depressuring Systems, fifth ed.
capacity evaluation. Cox, O.J., Weirick, M.L., 1980. Sizing safety relief valve inlet lines. Chem. Eng. Prog.
V76 (11), 51e54.
Friedel, L., Schmidt, J., 1993. Design of long safety valve vent line systems for gas
relief. J. Loss Prev. Process Ind. V6, 293e301.
Kandell, P., 1981. Program sizes pipe and flare manifolds for compressible flow.
Chem. Eng. V88 (13), 89e93.
4. Conclusion Kern, R., 1975. How to size piping and components as gas expands at flow condi-
tions. Chem. Eng. V82 (22), 125e132.
This paper proposes an easy and quick method to identify Lapple, C.E., 1943. Isothermal and adiabatic flow of compressible fluids. Trans. AIChE
V39, 385e432.
whether the venting line exit of the ERS with rupture disk is choked Levenspiel, O., 1977. The discharge of gases from a reservoir through a pipe. AIChE J.
or not. The relative error comparing to theoretical solution for this V23, 402e403.
method is less than 0.5% for any practical application. For capacity Van Boskirk, B.A., 1982. Sensitivity of relief valves to inlet and outlet line lengths.
Chem. Eng. V89 (17), 77e82.
evaluation of ERS, solving the adiabatic flow with friction is
Westman, M.A., 1997. Rapidly design safety relief valve inlet piping systems. Chem.
necessary to get Mach number and state variables at any location of Eng. Prog. V93 (3), 80e88.
venting line. For validating the adequacy of ERS, a quick and Westman, M.A., 1998. Quickly design subsonic safety relief systems. Chem. Eng.
straightforward method is proposed. This method circumvents Prog. V94 (2), 51e60.
White, F.M., 2001. Fluid Mechanics. McGraw-Hill.
solving adiabatic flow with friction. Three case studies are pre- Yaws, C.L., 2004. Yaws' Handbook of Thermodynamic and Physical Properties of
sented to support the proposed method. Chemical Compounds. Norwich, New York.
Please cite this article in press as: Zhao, G., An easy method to design gas/vapor relief system with rupture disk, Journal of Loss Prevention in the
Process Industries (2015), http://dx.doi.org/10.1016/j.jlp.2015.03.023