Professional Documents
Culture Documents
LNG FPSO
1
2,
1
2
3
Abstract
A layout of an LNG FPSO should be elaborately determined as compared with that of an onshore plant because many topside process
systems are installed on the limited area; the deck of the LNG FPSO. Especially, the layout should be made as multi-deck, not
single-deck and have a minimum area. In this study, a multi-floor layout for the liquefaction process, the dual mixed refrigerant(DMR)
cycle, of LNG FPSO was determined by using the optimization technique. For this, an optimization problem for the multi-floor layout was
mathematically formulated. The problem consists of 589 design variables representing the positions of topside process systems, 125
equality constraints and 2,315 inequality constraints representing limitations on the layout of them, and an objective function
representing the total layout cost. To solve the problem, a hybrid optimization method that consists of the genetic algorithm(GA) and
sequential quadratic programming(SQP) was used in this study. As a result, we can obtain a multi-floor layout for the liquefaction
process of the LNG FPSO which satisfies all constraints related to limitations on the layout.
Keywords : LNG FPSO( LNG , , ), Liquefaction process( ), Multi-floor layout( ), Optimization(
)
1.
1.1
LNG LNG-FPSO(Liquefied
(single-deck) (multi-deck)
Design) .
(turret) ,
(NG:
(separation process),
FPSO 70%,
process) ,
: miroh@ulsan.ac.kr, 052-259-2165
1.2
Penteado and Ciric (1996) (safety)
(single-floor) (chemical process plant)
.
.
,
(layout cost),
TNT (Trinitrotoluene equivalency
method)(Park, et al., 2011)
MILP
layout .
LNG FPSO
.
(heat exchanger)
. Patsiatzis and Papageorgiou
(2002)
, Penteado and Ciric (1996)
. Patsiazis (2002)
.
Table 1(a), (b) .
Object
Penteado
&
Ciric(1996)
EO plant1)
Patsiatzis
&
Papageorgi
ou
(2002)
Instant coffee
process plant,
EO plant
.
MINLP(Mixed Integer
Nonlinear Programming)(Lee & Leyffer, 2012)
layout .
Patsiatzis and Papageorgiou (2002)
.
, Penteado and Ciric(1996)
.
MILP(Mixed Integer Linear Programming)
layout .
Park (2011)
. Patsiatzis
and Papageorgiou (2002) ,
Patsiazis (2002) Penteado and
Ciric(1996) TNT equivalency method
.
MILP layout .
Georgiadis (1999)
. multi-floor
.
,
Patsiazis (2002)
(pump)
.
EO plant,
Park, et al. Benzene
production
(2011)
process plant
Instant coffee
Michael, et
process plant,
al.
Industrial
(1999)
multipurpose batch
plant
process
This study Liquefaction
of the LNG FPSO
Multi
-floor
Design
Variables
Position of each
equipment,
X
Land area, Safety
(Single devices
that have to
floor) be installed
at each
equipment
Floor allocation,
position and
orientation of each
equipment,
Land area
Floor allocation,
position and
orientation of each
equipment,
Land area
Floor allocation,
position and
orientation of each
equipment,
Land area
Floor allocation,
position and
orientation of each
equipment,
Land area
2. LNG FPSO
(cycle) (cascade
liquefaction cycle), (mixed refrigerant cycle),
(turbine-based cycle)
(Venkatarathnam, 2008),
DMR(Dual Mixed Refrigerant) LNG FPSO
.
69
LNG FPSO
Study
Additional Considerations
Equipme
nt
Safety consideration occupyin
g more
than
one floor
Minimum distance
between
equipment,
Installation of the1)
Penteado protection
devices
&
equipment cost,
Ciric(1996) for
Financial risk cost2)
predicted by the
TNT equivalency
method3)
Patsiatzis
&
Papageorgi
X
ou
(2002)
Explosion damage
Park, et al. cost predicted by
(2011) the TNT equivalency
method
Objective
Function
Optim
izatio
n
Meth
od
(compressor
suction drum), (overhead
crane) .
,
,
.
Layout cost4)
+ Protection
devices cost + MINL
5)
Financial risk P
cost
,
.
(precooling part)
precooling mixed refrigerant(PMR) receiver .
Layout cost
6
MILP
)
2~3
(buffer tank)
Michael, et
al.
(1999)
This study
Minimum distance
between
equipment
Layout cost
+ Explosion MILP
damage cost
surge
.
.
Layout cost +
Upward and
horizontal
MILP
transportation
cost7)
Hybri
d
optimi
Layout cost zation
meth
od
(SQP
+GA)
70
49 1 2012 2
DMR
, 3 (impeller)
3 1, 1,
1, 3, (SW
cooler) 3, PMR receiver 1, 3, (valve) 3
, (liquefaction part) (subcooling part)
2 2 1,
1, 1, 2,
2, 1, 2 MR
(mixed refrigerant separator) .
Table 2 .
(module) (Mecklenburgh,
1985),
. DMR
Fig. 1 3 .
3
3 1,
1, 1, 3,
3 , PMR1
.
PMR receiver 1, 3
, 3 , PMR2
. ,
2
1, 1,
1, 2, 2,
1, 2 MR ,
MR .
,
8m . PMR2
MR A E ,
PMR1 A D .
MR
. , Fig. 2
(maintenance area) y
9m ( ),
. A
MR
(working space)
50%
( ).
E
(safety facilities)
60%
. 4m ,
ai [m]
bi [m]
hi [m]
13
43
4.5
4.5
13
12
12
12
19
1
1
1
1
1
1
71
LNG FPSO
3m
. ,
.
MR
.
3. Liquefaction process
3.1 (design variables)
,
, (Patsiatzis & Papageorgiou,
2002).
(continuous variables)
xi,yi: i x, y
zi: i i
Ri,j: x i j i j
x
Li,j: x i j i j
x
Ai,j: y i j i j
y
Bi,j: y i j i
j y
Ui,j: i j i
j y
Di,j: i j i
j y
TDi,j: i j (total rectilinear
distance)
FA:
Xmax: x
Ymax: y
(binary variables)
Vik: i k 1, 0
Zij: i j 1,
0
Oi: i ai bi ai x (, ai
li) 1, 0
E1i,j, E2i,j: (non-overlapping
constraints)
i, j , k
72
Equipment
xi
No. Name
[m]
MR
separator
1
x1
on lower
deck
MR
separator
2
x2
on upper
deck
MCHE
3
x3
on A deck
MCHE
4
x4
on B deck
MCHE
5
x5
on C deck
MCHE
6
x6
on D deck
MCHE
7
x7
on E deck
MR Comp.
suction
8 drum on
x8
lower
deck
MR Comp.
suction
9 drum on
x9
upper
deck
10 MR Comp. x10
Cooler for
11
x11
comp.
Overhead
12
x12
crane
SW water
13
x13
4
SW water
14
x14
5
15 Valve 4
x15
16 Valve 5
x16
yi
[m]
Oi
y1
Vi,1
Vi,k
...
Vi,5
O1
V1,1
...
V1,5
y2
O2
V2,1
...
V2,5
y3
O3
V3,1
...
V3,5
y4
O4
V4,1
...
V4,5
y5
O5
V5,1
...
V5,5
y6
O6
V6,1
V6,5
y7
O7
V7,1
V7,5
y8
O8
V8,1
V8,5
y9
O9
V9,1
V9,5
y10
O10
V10,1
V10,5
y11
O11
V11,1
V11,5
y12
O12
V12,1
V12,5
y13
O13
V13,1
V13,5
y14
O14
V14,1
V14,5
y15
y16
O15
O16
V15,1
V16,1
V15,5
V16,5
(FA), x
(Xmax), y (Ymax)
589.
3.2 (constraints)
3.2.1 (equality constraints)
(1) (deck constraints)
,
49 1 2012 2
, i = 1, 2, , 16
li: (x )
From
equipment i
to
equipment j
1 to 4
4 to 15
15 to 4
2 to 5
5 to 6
6 to 16
16 to 7
3 to 8
9 to 10
10 to 13
13 to 10
10 to 14
10 to 11
11 to 10
di: (y )
Ri,j
[m]
Li,j
[m]
Ai,j
[m]
Bi,j
[m]
Ui,j
[m]
Di,j
[m]
TDi,j
[m]
R1,4
R4,15
R15,4
R2,5
R5,6
R6,16
R16,7
R3,8
R9,10
R10,13
R13,10
R10,14
R10,11
R11,10
L1,4
L4,15
L15,4
L2,5
L5,6
L6,16
L16,7
L3,8
L9,10
L10,13
L13,10
L10,14
L10,11
L11,10
A1,4
A4,15
A15,4
A2,5
A5,6
A6,16
A16,7
A3,8
A9,10
A10,13
A13,10
A10,14
A10,11
A11,10
B1,4
B4,15
B15,4
B2,5
B5,6
B6,16
B16,7
B3,8
B9,10
B10,13
B13,10
B10,14
B10,11
B11,10
U1,4
U4,15
U15,4
U2,5
U5,6
U6,16
U16,7
U3,8
U9,10
U10,13
U13,10
U10,14
U10,11
U11,10
D1,4
D4,15
D15,4
D2,5
D5,6
D6,16
D16,7
D3,8
D9,10
D10,13
D13,10
D10,14
D10,11
D11,10
TD1,4
TD4,15
TD15,4
TD2,5
TD5,6
TD6,16
TD16,7
TD3,8
TD9,10
TD10,13
TD13,10
TD10,14
TD10,11
TD11,10
(4)
(5)
, i j
MCHE(Main Cryogenic Heat Exchanger)
,
,
zi
.
Equipment j
16
Equipment i
1
Z1,2,E11,2,E21
Z1,16,E11,16,E21,16
Z2,16,E12,16,E2216
Z15,16,E115,16,E215,
,2
15
16
16
(6)
, i j
NF: Number of deck(=5)
H: Height between decks(=8m)
Table 6 Value of zi
(1)
i j ( i
j , total rectilinear distance)
(7)
Oi
(Patsiatzis & Papageorgiou, 2002),
32
, i j
56
.
(2)
(3)
(FA) x
73
LNG FPSO
max
max
(X ) y (Y ) ,
3.2.2 (4)
(additional layout design constraints) .
y Fig. 2
y 9m
.
m ax m ax
(8)
(17)
(18)
.
MR 2 (19)
.
(19)
MCHE 5 (20)
.
(20)
, k = 1
2 (21)
.
(9)
(10)
(21)
MR , ,
(22) .
(11)
(12)
(22)
20 .
, i = 3, 4, 5, 6
2
(13), (14) .
(13)
(14)
Zij
MR MR
, MR
, 1,800(= )
i j
.
x y
(23)
(24)
(25)
(15)
(16)
74
49 1 2012 2
for non-overlapping .
(30)~(33)
(Zij = 1), x y
4m , (30)~(33)
. Zij = 1 , 4
(34)
(35)
(26)
(27)
(35)
(28)
(37)
(29)
, i = 1, 2, , 15, j = i+1, , 16
480(= )
, i = 1, 2, , 15, j = i+1, , 16
(26)~(29) Zij = 1
,
.
(Patsiatzis & Papageorgiou, 2002).
(30)
(31)
,
2,280(=1,800+480) .
(2) (working space area constraints)
A MR
50% ,
.
(32)
(33)
, i = 1, 2, , 15, j = i+1, , 16
, M
,
(38)
(39)
. E1ij E2ij
(26)~(29) . , j > i i, j
(30) . (27) ,
60%
, (32) .
(29) , E1ij = 1, E2ij = 1 ,
,
.
(40)
75
LNG FPSO
fij: i j 1,
0
c
C ij: i j
Fig. 2
3m
.
.
Cvij: i j
Chij: i j
FC :
(41)
LC :
(42)
MR
. ,
, i = 1, 2, , 16
32 .
, Fig. 2
3m
10 .
. x y
. ,
m ax
(43)
m ax
(44)
(46)
3.4
([]
).
, i = 1, 2, , 16
: Minimize W (X)
32 .
:
( ) [125]
- [16]
- [32]
- [56]
(layout cost)
- [1]
- [20]
2011).
( ) [2,315]
- [2,280]
- [2]
- [1]
- [32]
(45)
, X = {xi, yi, zi, Ri,j, Li,j, Ai,j, Bi,j, Ui,j, Di,j, TDi,j, FA,
Xmax, Ymax, Vik, Zij, Oi, E1i,j,E2i,j}.
, LNG FPSO MR
589 , 125 2,315
, i = 1, 2, , 16, j = 1, 2, , 16
76
49 1 2012 2
4.
LNG FPSO
. GA(Genetic
Algorithm) SQP(Sequential Quadratic Programming)
hybrid ,
EzOptimizer(Lee, et al., 2002)
.
. Table 7
,
Fig. 3 .
Rij, Lij, Aij, Bij, Uij, Dij, TDij, Zij, E1ij, E2ij Table 7
, .
3.2
. Table 7 Fig. 3
3.2.2 (1)
.
yi
xi
[m] [m]
Oi
12
Name
MR
separator
1
20
on lower
deck
MR
separator
2
20
on upper
deck
MCHE
3
20
on A deck
MCHE
4
20
on B deck
MCHE
5
20
on C deck
MCHE
6
20
on D deck
MCHE
7
20
on E deck
MR Comp.
suction
8
20
drum on
lower deck
MR Comp.
suction
9
20
drum on
upper deck
10 MR Comp. 10
Cooler for
11
10
comp.
Overhead
12
10
crane
13 SW water 4 21
14 SW water 5 6
15 Valve 4 15
16 Valve 5 14
26.0
max
max
Y
X
m
Vi,k
Vi,1
Vi,2
Vi,3
Vi,4 Vi,5
12
20
20
11
11
11
11
11
7
7
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
1
0
1
31.8 m
FA
5.
826.8 m
LNG FPSO
77
LNG FPSO
DMR MR
. ,
,
( ) .
GA SQP hybrid
,
.
,
.
LNG FPSO ,
.
TNT
,
.
(a) (KRF-2008-314-D00494,
KRF-2009-0086033, R33-2008-000-10150-0) (b)
SM-11 /
(Virtual) (Architecture) (c)
(10035331)
(d) BK21
(e) (f)
GET-Future (No. 20114030200050)
, .
78
49 1 2012 2