Energy xxx (2015) 1e13

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Application of metal foam heat exchangers for a high-performance

liquefied natural gas regasification system
Dae Yeon Kim, Tae Hong Sung, Kyung Chun Kim*
School of Mechanical Engineering, Pusan National University, Busan 609-735, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: The intermediate fluid vaporizer has wide applications in the regasification of LNG (liquefied natural gas).
Received 20 January 2015 The heat exchanger performance is one of the main contributors to the thermodynamic and cost
Received in revised form effectiveness of the entire LNG regasification system. Within the paper, the authors discuss a new
5 October 2015
concept for a compact heat exchanger with a micro-cellular structure medium to minimize volume and
Accepted 17 October 2015
mass and to increase thermal efficiency. Numerical calculations have been conducted to design a metal-
Available online xxx
foam filled plate heat exchanger and a shell-and-tube heat exchanger using published experimental
correlations. The geometry of both heat exchangers was optimized using the conditions of thermolators
Keywords:
Metal foam heat exchanger
in LNG regasification systems. The heat transfer and pressure drop performance was predicted to
LNG (liquefied natural gas) compare the heat exchangers. The results show that the metal-foam plate heat exchanger has the best
Regasification system performance at different channel heights and mass flow rates of fluid. In the optimized configurations,
the metal-foam plate heat exchanger has a higher heat transfer rate and lower pressure drop than the
shell-and-tube heat exchanger as the mass flow rate of natural gas is increased.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction
Gas heating with ambient heat sources is required in various
cryogenic applications. After extracting cold exergy from cryogenic
sources, the temperatures of the source fluids are still below zero
degrees Celsius since specific applications do not fully utilize all
exergy in cold sources. This low temperature could cause several
technical problems. In the application of cold exergy in LNG (lique-
fied natural gas), solid methane hydrate could be generated at
temperature below zero and adhere to pipe walls. As a result,
pipeline pressure losses are increased, and in the worst case, the pipe
is blocked with methane clathrate. Similarly, cold fluid could cool
down the pipe to below zero, and frost layers could be formed
outside the pipe. During pipeline transport, there are natural tem-
perature drops from pressure losses caused by friction. In this case,
the fluid is preheated to more than zero degrees Celsius before
entering the local pipeline network to compensate for the temper-
ature drops caused by the JouleeThompson effect. To prevent these
problems, a gas-heater is used in cryogenic applications [1e11].
Many have investigated the heat and mass transfer character-
istics in the medium of compact heat exchangers, such as wire
screen and metal-foam structures. Tian et al. [12] researched fluid
flow and heat transfer of cellular metal lattice structures. Paek et al.
[13] and Calmidi et al. [14,15] presented a comprehensive analytical
and experimental investigation for determination of the effective
thermal conductivity (ke), permeability (K) and inertial coefficient
(f) of high-porosity metal foams. Liu et al. [16] measured the
pressure drop through various types of foam matrixes and devel-
oped empirical equations for the friction characteristics of a foam
matrix. Lu et al. [17] presented an analytical study of the forced
convection heat transfer characteristics in high-porosity open-cell
metal foam-filled pipes. Tadrist et al. [18], Vafai et al. [19] and Kim
et al. [20,21] investigated the impact of porous fins on the pressure
drop and heat transfer characteristics in plate-fin heat exchangers.
This study presents a novel concept of a metal-foam plate heat
exchanger for weight reduction and higher thermal efficiency.
Based on empirical data and correlations of heat and mass transfer,
heat exchangers were designed for the thermolators in LNG rega-
sification systems. The exchangers were also compared to con-
ventional shell-and-tube heat exchangers.
2. Physical model
2.1. LNG regasification systems

* Corresponding author. Tel.: þ82 51 510 2324; fax: þ82 51 512 9835. The IFV (intermediate fluid vaporizer) is a new kind of vaporizer
E-mail address: kckim@pusan.ac.kr (K.C. Kim). for LNG regasification systems. A thermal model has been

http://dx.doi.org/10.1016/j.energy.2015.10.056
0360-5442/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
2 D.Y. Kim et al. / Energy xxx (2015) 1e13

Table 2
Dimensions of shell-and-tube heat exchanger.

Parameters Dimension

Tube-side outside diameter, do(mm) 20

Tube-side inside diameter, di (mm) 16
Tube pitch, pt (mm) 25
Tube bundle layout ( ) 45
Central baffle spacing, Lb,c (m) 0.279
Outlet baffle spacing, Lb,o (m) 0.318
Inlet baffle spacing, Lb,i (m) 0.318
Tube length, Lt (m) 2e11
Shell side inside diameter, Ds (m) 0.78e1.78
Baffle cut, lc (mm) 86.7
Number of sealing strip pairs, Nss 1
Width of bypass lane, wp (mm) 19.0
Fig. 1. Schematic of IFV with thermolator. Number of tube passes, np 2
Number of pass partitions, Np 2
Tube-to-baffle hole diametral clearance, dtb (mm) 0.794
Table 1 Shell-to-baffle diametral clearance, dsb (mm) 2.946
Default values of the known parameters. Heat Transfer Area (m2) 470
Total Volume (m3) 4.98e5.24
TCold_in (K) THot_in (K) PCold_in (bar) PHot_in (bar) m_ Cold in (kg/s) m_ Hot in (kg/s)

233 288 122 10 50 1000

rotated-square arrangement. The geometric parameters are shown

developed on the energy balance of the three typical parts of IFV:
the evaporator, condenser, and thermolator. A schematic of the
system is shown in Fig. 1.
LNG vaporization is achieved using two main heat exchangers
operating in series. The dotted box in Fig. 1 is a second medium
cycle that is driven using the latent heat of propane condensation to
partially heat the LNG. The IF (intermediate fluid) transfers its
latent heat during vaporization and condensation using seawater as
a heat source and LNG as a cooling source. The thermolator using
seawater heats LNG further to the required temperature. LNG is
pumped into the condenser by a pump, where it gasifies. The
natural gas from the condenser is heated continuously in the
thermolator by the seawater, and then goes into the gas-supplying
system. Since LNG is pre-heated by a second fluid, it avoids direct
contact with seawater and cryogenic LNG and prevents freezing of
the seawater. For this reason, seawater close to freezing can be used
in this configuration [22e26]. The default values of known fluid
parameters are shown in Table 1.
2.2. Shell-and-tube heat exchangers
Shell-and-tube heat exchangers have many different configu-
rations. This paper investigates the TEMA E type exchanger with
one shell and two tube passes. Seawater flows on the tube side and
the vaporized LNG flows on the shell side. The models for the
pressure drop and heat transfer correlations are thus considered as
a convective single-phase flow. The tubes in the bundle are in a 45
in Fig. 2.
The dimensions of the exchanger are shown in Table 2. To find
the optimized geometry of the shell with the best heat and mass
transfer performance, a simulation was conducted with different
aspect ratios of the tube length and shell-side inner diameter with a
fixed total heat transfer area.
2.3. Metal-foam plate heat exchangers
Plate heat exchangers can have different types of corrugations
[27], such as chevron or herringbone corrugations. However, in this
study, the inner shape of the plate heat exchanger is aluminum
metal-foam instead of general corrugations. The conceptual design
and geometrical parameters of the metal-foam plate heat
exchanger are shown in Fig. 3. A one-pass plate heat exchanger
with a channel filled with metal-foam is used to heat cryogenic NG
with seawater on the other side.
To find the optimized plate configuration with the best heat and
mass transfer performance, a simulation was conducted with
different channel heights at a given total volume. Table 3 shows
dimensions of the metal-foam plate.
The inner structure of metal-foam is shown in Fig. 4. Specimens
with porosity of ε ¼ 0.92 and pore density of 10PPI (pore per inch)
were selected so that the porosity would be comparable to that of
conventional fins used in heat exchangers. The permeability of the
aluminum foams was determined from pressure drop data
measured by Kim et al. [20,21]. The metal-foam parameters are
shown in Table 4. The fiber diameter, df was 0.2 mm and the pore
size, dp was 1 mm.

Fig. 2. Geometrical characteristics of a shell-and-tube heat exchanger. Figure adopted from Ref. [27].

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
 D.Y. Kim et al. / Energy xxx (2015) 1e13 3

Fig. 3. Geometrical parameters of metal-foam plate heat exchanger.

Table 3
Geometric dimensions of metal-foam plate heat exchanger. f ¼ A þ B$Re1=m ðBhatti  shah correlationÞ
Parameters Dimension A ¼ 0:0054; B ¼ 2:3  108 ; m ¼ 2=3 2100  Re  4000
Port diameter Dp (m) 0.3
A ¼ 0:00128; B ¼ 0:1143  108 ; m ¼ 3:2154 4000  Re  107
Channel width W (m) 0.75 (1)
Channel height b (mm) 5e150
Thickness of plate t (mm) 2 The Fanning friction factor is used to obtain the total pressure
Length of a plate between ports Lp (m) 2.1 drop on the tube side:
Length of a plate for heat transfer Lh (m) 1.5
Heat Transfer Area (m2)

4fL   
27e1400
Total Volume (m3) m_ 2t 2 2
5.05 DPt ¼ þ 1  s þ Kc  1  s  Ke np (2)
2gc rt A2o;t di

where s ¼ 2(pt  do)/2pt, Ke ¼ (1  s)2 is the sudden expansion loss

factor, and Kc ¼ 0.5(1  s) is the sudden contraction loss factor
generated at the entrance of the heat exchanger. The fluid experi-
ences a sudden contraction and expansion while flowing from the
manifold to the tube so these effects should be considered.

3.1.2. Convective heat transfer correlations for seawater (tube side)

The Gnielinski correlation from Ref. [27] was applied to the
convective heat transfer coefficients on the tube side. The formula
is expressed in Eq. (3). The characteristic length of the correlation is
the inner diameter of the tube and the fanning factor f is obtained
from the Bhattieshah correlation in Eq. (1).
Fig. 4. SEM (scanning electron microscope) image of metal foam with 20ppi porosity.
ðf =2ÞðRe  1000ÞPr
Nu ¼   (3)
3. Numerical model 1 þ 12:7ðf =2Þ1=2 Pr 2=3  1

3.1. Shell-and-tube heat exchangers where 2300  Re  5  106, 0.5  Pr  2000.

3.1.1. Convective pressure drop correlations for seawater (tube side)
The Bhattieshah correlation was used for the pressure drop
prediction on the tube side and is presented in Eq. (1). The Fanning
friction factor depends on the Reynolds number and flow cross-
section geometry.
3.1.3. Convective pressure drop correlations for natural gas (shell
side)
The BelleDelaware method [27] is used to calculate the pressure
drop and heat transfer coefficient on the shell side. The total
pressure drop on the shell side is the sum of the pressure drop
associated with each cross flow section between baffles, the

Table 4
Flow parameters of Al-6101 aluminum foam.

Pore per inch (PPI) Porosity ε Fiber diameter df (mm) Pore diameter dp (mm) Permeability K (m2) Surface area to volume ratio (m2/m3)

10 0.92 0.2 1 1.04 790

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
4 D.Y. Kim et al. / Energy xxx (2015) 1e13

pressure drop associated with each window section, and the outlet sections, respectively. Dpb,id and Dpw,id are the ideal frictional
pressure drop for cross flow sections on each end between the first pressure drops in the cross flow section and window section,
and last baffle and the tube sheet. The shell-side frictional pressure respectively. These pressure drops should be corrected by z
drop is given as follows: correction factors, as shown in Table 5. zb, zl and zs are the

h i 
Nr;cw
Dps ¼ Dpcr þ Dpw þ Dpio ¼ ðNb  1ÞDpb;id zb þ Nb Dpw;id zl þ 2Dpb;id 1 þ z z
Nr;cc b s

4f G2 Nr;cc mw 0:25
Dpb;id ¼ id s
2gc rs mm (4)
  r u2
Dpw;id ¼ 2 þ 0:6Nr;cw m z
2gc
fi;d ¼ 3:5ð1:33do =pt Þb $Re0:476
s

where b ¼ 6:59=ð1 þ 0:14Re0:52

s Þ correction factors for the bypass flow, baffle leakage, and larger
baffle spacing at the inlet and outlet, respectively.
Dpcr ¼ DPb;id ðNb  1Þzb zl
3.1.4. Convective heat transfer correlations for natural gas (shell
side)
  G2w The convective heat transfer coefficient on the shell side of the
DPw ¼ Nb 2 þ 0:6Nr;cw z shell-and-tube heat exchanger was obtained by the Zhukauskas
2gc rs l
correlation presented in Eq. (5). The characteristic length is the
outer diameter of the tube.

 . 0:25
8 1:04Re0:4 0:36
1 < Re < 102
> f Prf Prf Prow
>
>
>
>  . 0:25
>
>
>
> 0:71Re0:5 0:36
5  102 < Re < 103
>
> f Prf Prf Prow
>
>
>
< s1 0:2 0:6 0:36  . 0:25 s1
Nuf ¼ 0:35 Ref Prf Prf Prow  2; 103 < Re < 2  105 (5)
>
> s2 s2
>
>  . 0:25
>
> s1
>
> 0:40Re0:6 0:36
> 2; 103 < Re < 2  105
>
> f Prf Prf Prow
>
> s2
>
>
: s1 0:2 0:8 0:36  . 0:25
0:031 Ref Prf Prf Prow 2  105 < Re < 2  106
s2

 where the subscript ow represents the value of the properties

Nr;cw
Dpio ¼ 2Dpb;id 1 þ z z evaluated at the tube outer wall temperature:
Nr;cc b s

where Dpcr, Dpw and Dpio are the frictional pressure drops in the Tm;h hh þ Tm;c hc
cross flow section, window area section, and shell-side inlet and Tow ¼ (6)
hh þ hc
The J correction factors for non-idealities corrects the ideal heat
Table 5 transfer coefficient obtained for cross flow over a tube bundle from
z correction factors for the pressure drops on the shell side. Eq. (5). The shell-side heat transfer coefficient is given as Eq. (7) and
z Correction factors, Formula for correction factor the correction factors are shown in Table 6.
parameters
8 Ao;bp
> 1 rb ¼
> þ 1=3 þ
< expf  Drb ½1  ð2Nss Þ g for Nss < Ao;cr
þ ¼ Nss
hs ¼ hid Jc Jl Jb Js Jr (7)
2 Nss
zb ¼ N
>
> 1
r;cc
: þ
1 for Nss  4:5 for Res  100 The heat transfer rate formulated by the heat transfer equation
2 D¼
3:7 for Res > 100 for the shell-and-tube heat exchanger is presented in Eq. (8)
p
zl ¼ exp½1:33ð1 þ rs Þrlm  P ¼ [0.15(1 þ rs) þ 0.8]
!2n0 !2n0

1:0 for laminar flow

Lb;c Lb;c n0 ¼
zs ¼ Lb;o þ Lb;i 0:2 for turbulent flow QSTHE ¼ USTHE ASTHE DTlm (8)
STHE

Source: Data from Taborek (1998) [27].

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
 D.Y. Kim et al. / Energy xxx (2015) 1e13 5

Table 6
J correction factors for the heat transfer on shell side.

J Correction factors Formula for parameters for correction factors

ctl Þ
Jc ¼ 0.55 þ 0.72Fc Fc ¼ 1  2Fw ¼ 1  qpctl þ sinðq
p
Jl ¼ 0:44ð1  rs Þ þ ½1  0:44ð1  rs Þe2:2rlm A
rs ¼ Ao;sb þA
o;sb A þA
; rlm ¼ o;sbAo;cr o;tb

o;tb 
Ao;cr ¼ Ds  Dotl þ 2 DXctlt ½ðpt  do Þ  ðde  do ÞdNf  Lb;c

Ao;tb ¼ p4 ½ðdo þ dtb Þ2  d2o Nt ð1  Fw Þzpdo dtb N2t ð1Fw Þ


( Ao;sb ¼ pDs d2sb 1  2p qb
þ
1 for Nss  1=2
Jb ¼ þ 1=3
Crb ½1ð2Nss Þ  þ
A
rb ¼ Ao;bp þ ¼ Nss
, Nss
e for Nss  1=2
o;cr
Nr;cc
1:35 for Res  100
C¼
1:25 for Res > 100
Nb 1þðLþ Þð1nÞ þðLþo Þ
ð1nÞ
Js ¼ i
Lþ
L
¼ Lb;i , Lþ
L
o ¼ Lbc
b;o
Nb 1þLþ i
þLþ
o i
8 bc
< 0:6 for turbulent flow
n¼
: 1 for laminar flow

3
1 for Res  100
Jr ¼ Nr,c ¼ Nr,cc þ Nr,cw
ð10=Nr;c Þ0:18 for Res  20

Nr;cw ¼ 0:8 Xl lc  2 ðDs  Dctl Þ
1

Nr;cc ¼ Ds 2l
Xl
c

Source: Data from Taborek (1998) [27].

where USTHE is the overall heat transfer coefficient and DTlm_STHE is 3.2.2. Convective heat transfer correlations
the logarithmic mean temperature for the shell-and-tube heat A heat transfer correlation of a metal-foam-filled fin plate was
exchanger, which are determined using Eq. (9) and Eq. (10) proposed by Kim et al. [21], which was applied to both sides of the
  metal-foam plate heat exchanger:
1 1 Ao do do 1
¼ þ ln þ (9)
USTHE hi Ai 2kwall di ho

ðThm  Tco Þ  ðThm  Tci Þ

DTlm STHE ¼  (10)
Tco
ln TThm
hm Tci

3.2. Metal-foam plate heat exchangers

3.2.1. Convective pressure drop correlations

The correlation of friction factor f for the aluminum foams
developed by Kim et al. [21] is adopted for the pressure drop on
both sides of the metal-foam plate heat exchanger:

1 CE
f¼ þ (11)
ReDa Da1=2

where, Re ¼ Uyi H; and CE ¼ 0.1 is the inertia coefficient for the

present aluminum foams. It should be noted that the equation uses
the Forchheimer-extended Darcy model for porous media, which
includes the inertia effect [28]. The Darcy number is Da ¼ bK2 ; where
K is the permeability and b is the channel height.
For the vertical upward flow of a plate heat exchanger, the total
pressure drop consists of the pressure drop associated with the
inlet and outlet manifolds and ports, the frictional pressure drop
within the core, the accelerational pressure drop, and the static
head loss [27], as expressed in Eq. (12).

DPtotal ¼ DPport þ DPfr þ DPacc þ DPhead

 
1:5G2p np 4fLG2 1 1 1 G2 rmgL
¼ þ þ  ± (12)
2gc ri 2gc De r m r0 ri gc gc
Fig. 5. Schematic of the solution procedure for heat exchangers.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
6 D.Y. Kim et al. / Energy xxx (2015) 1e13

QMFHE ¼ UMFHE AMFHE DTlm MFHE (15)

where UMFHE is the overall heat transfer coefficient and DTlm_MFHE is

the logarithmic mean temperature for the metal-foam plate heat
exchanger:

1 1 d 1
¼ þ þ (16)
UMFHE hhot kwall hcold

ðThi  Tco Þ  ðTho  Tci Þ

DTlm ¼  (17)
ln TThihoTco
Tci

3.3. Solution procedure

The solution procedure for the heat exchanger models is sche-

Fig. 6. Pressure losses on the tube-side according to the ratio of the tube length to the
shell diameter for different mass flow rates of cold fluid. matically shown in Fig. 5. After inputting the design specifications
mc, mh, Tci, Thi, Pci, and Phi, the log mean temperature difference DTlm
and Tout are assumed based on the input values of the initial fluids
properties and the geometry of the heat exchangers at an initial
Nu ¼ 0:0159Re0:426 Pr 1=3 Da0:787 (13) effectiveness 0.5. Then, the pressure drop and heat transfer per-
formance of the designed heat exchanger models can be calculated
Generally, the heat transfer coefficient can be derived from the with the correlations and using the εNTU method in a diverse
Nusselt number and the thermal conductivity of the fluids. How- range of mass flow rates. NTU and the effectiveness can then be
ever, due to the complex and porous structure of the metal-foam, obtained, and all of the processes are iterated until the error of the
the effects of the heat conduction in the solid part of the metal- effectiveness is lower than the tolerance of 0.0001%. Finally, the
foam with the fluid should be considered. It is necessary to use overall heat transfer coefficient, pressure drop, heat transfer rate,
the effective thermal conductivity for the heat transfer coefficient DTlm, and Tout can be obtained.
in the metal-foam channel. The correlation of the effective thermal In the thermolators of the LNG regasification systems, LNG or
conductivity in the metal-foam channel flow developed by Calmidi NG is used as a refrigerant, and seawater is used as a heat source.
et al. [14] is adopted: Methane was used as a working fluid on the cold side because LNG
and NG are mostly composed of methane. The salt solution in the
  0:65
ke ¼ 0:35 εkf þ ð1  εÞks þ ! (14) seawater was ignored for the heat transfer process. Therefore, the
ε
thermal properties of water and methane were used for the
kf
þ 1ε
k working fluids and were obtained from code based on REFPROP
s

[29], which is developed by the National Bureau of Standards. The

The heat transfer rate formulated by the heat transfer equation calculation process was conducted using MATLAB [30] code
for the metal-foam plate heat exchanger is presented in Eq. (15): therefore the REFPROP-MATLAB code was used.

Fig. 7. Pressure losses on the shell-side according to the ratio of the tube length to the shell diameter for different mass flow rates of cold fluid.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
 D.Y. Kim et al. / Energy xxx (2015) 1e13 7

Fig. 8. Hot-side outlet temperature according to the ratio of the tube length to the shell diameter with different cold fluid mass flow rates.

4. Results and discussion
4.1. Effects of heat exchanger geometry
The thermo-fluidic performance of the heat exchanger varies
with the operation and structural conditions. The state of the
product fluid also changes according to the conditions of the
introduced fluid. The performance of the heat exchanger varies
with the shape as well, which is highly dependent on the boundary
layer. For example, in the shell-and-tube heat exchanger, the
required product fluid conditions are satisfied with a specific ratio
of the shell diameter to the tube length, tube size, tube configura-
tion, etc. We analyzed the performance of the heat exchanger
according to the geometry and we compared the thermo-fluidic
performance of the different types of exchangers. For the compar-
ison, the total heat transfer area and heat source were kept the
same in the analysis.
4.1.1. Shell-and-tube heat exchanger
Figs. 6e12 show the effects of the ratio of the tube length Ltube to
the shell inner diameter Ds, and the mass flow rate of cold flow
(vaporized NG) for the shell-and-tube heat exchanger. The ratio
corresponds to the shape of the heat exchanger, and a larger ratio
indicates an elongated appearance of the heat exchanger. For the
same total heat exchange area, a larger ratio means a smaller

Fig. 9. Cold-side outlet temperature according to the ratio of the tube length to the shell diameter with different cold fluid mass flow rates.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
8 D.Y. Kim et al. / Energy xxx (2015) 1e13

Fig. 10. LMTD according to the ratio of the tube length to the shell diameter with different cold fluid mass flow rates.

number of tube bundles and longer tube length, which causes a
higher pressure drop. The range of the ratio Ltube/Ds is 1e14.
The results are shown in Figs. 6 and 7 for the hot and cold sides.
In Fig. 6 (hot-side of the tube), as the ratio increases, the pressure
drop increases exponentially (Eq. (2)) from 0.26 bar to 25 bar. The
pressure drop also increases as the cold flow rate increases, but the
effect is relatively small. In Fig. 7 (cold-side of the shell), the pres-
sure drop also increases as the ratio increases, but the tendency is
different (Eq. (4)). In this case, the pressure drop increases expo-
nentially as the flow rate of the cold side increases. The pressure
loss of the hot side of the tube appears to be larger.
In Figs. 8 and 9, the outlet temperatures of both sides of the heat
exchanger are shown. When the cold-source mass flow rate is low,
both outlet temperatures approach the sea water inlet temperature
of 15  C. When the cold source flow rate is increased, both outlet
temperatures decrease. When the ratio is low, the degree of
temperature drop is also low. Therefore, in order to gain enough
heat exchange, sufficient tube length is required. In addition, as the
ratio increases, LMTD decreases (Fig. 10). The overall heat transfer
coefficient increases as the mass flow rate and the ratio increase
(Fig. 11). The overall heat transfer coefficient increased from
2600 W/m2K to 4500 W/m2K at a mass flow rate of 250 kg/s.
Similarly, the heat transfer rate increase with the ratio and mass
flow rate (Fig. 12). The highest heat transfer rate is 43 MW at a mass
flow rate of 250 kg/s.
4.1.2. Metal-foam plate heat exchanger
The effect of the channel height with a fixed total volume of the
heat exchanger was analyzed in the case of metal-foam plate heat
exchanger, and the results are shown in Figs. 13e20. The range of
the channel-height is 1e150 mm. Fig. 13 shows the total heat
transfer area due to the geometrical effects. As the channel height

Fig. 11. Heat transfer coefficient according to the ratio of the tube length to the shell diameter with different cold fluid mass flow rates.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
 D.Y. Kim et al. / Energy xxx (2015) 1e13 9

Fig. 12. The amount of heat transfer according to the ratio of the tube length to the shell diameter with different cold fluid mass flow rates.

Fig. 15. Pressure losses on cold-side according to the channel height with different
Fig. 13. Total heat transfer area according to the channel height. cold fluid mass flows.

Fig. 14. Pressure losses on hot-side according to the channel height with different cold Fig. 16. Outlet temperature of hot-side according to the channel height with different
fluid mass flows. cold fluid mass flows.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
10 D.Y. Kim et al. / Energy xxx (2015) 1e13

Fig. 17. Outlet temperature of cold side according to the channel height with different cold fluid mass flows.

increases with a fixed total volume, the total heat transfer area is
reduced because of the decreased number of plates.
Fig. 14 shows that the pressure losses on the hot side (where
seawater flows) depend on secondary pressure losses of the ports,
not the channel height. When the channel height increases, the
port pressure loss becomes relatively dominant over the frictional
loss in the core section, so the total pressure drop converges at
higher channel height of the metal-foam plate heat exchanger.
Similar results for the cold side are shown in Fig. 15. The pressure
losses are large in the low channel height range since the friction
losses are higher than the secondary losses in this range. The
pressure losses also increase as the cold source mass flow rate
increases.
Figs. 16 and 17 show the temperature change of each product
flow. In Fig. 16, specific channel heights for the minimum seawater
outlet temperature appear where the flow area is too small for the
low channel height. A higher cold source flow rate produces the
lower outlet temperatures (Figs. 16 and 17). LMTD also mono-
tonically increases as the channel height increases, as shown in
Fig. 18. Fig. 19 shows that the channel height and the cold source
mass flow rate have no significant effect on the heat transfer
coefficient.
Fig. 20 shows the optimum channel height for the maximum
amount of heat transfer. The total heat transfer rate is affected by
the overall heat transfer coefficient, log mean temperature differ-
ence, and total heat transfer area, as shown in Eq. (15). However,
the total heat transfer area decreases with increasing channel
height, as shown in Fig. 13, although the other parameters are
increased. This produces the optimal channel height. The higher
the flow rate, the thinner the channel height should be.

Fig. 18. LMTD according to the channel height with different cold fluid mass flows.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
 D.Y. Kim et al. / Energy xxx (2015) 1e13
Fig. 19. Heat transfer coefficient according to the channel height with different cold
fluid mass flows.
Fig. 20. Amount of heat transfer according to the channel height with different cold
fluid mass flows.
Fig. 21. Cold-side pressure drop according to cold source mass flow rate.
Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
11
Fig. 22. Hot-side outlet temperature according to cold source mass flow rate.
4.2. Effects of mass flow rate
The effects of the cold source mass flow rate were simulated.
The heat exchanger geometries at the same maximum pressure loss
of 6.0 bar (Figs. 6 and 14) were selected for comparative studies. In
this case, the metal-foam plate heat exchanger has channel height
10 mm, while for the shell-and-tube heat exchanger, the ratio of the
tube-length to the shell-diameter is 6.5 based on the results in
chapter 4.1.
In Fig. 21, the pressure drop on the cold side exponentially in-
creases as the cold source mass flow rate increases. The pressure
drop of the metal-foam plate heat exchanger is 0.041e2.249 bar,
and that of the shell-and-tube heat exchanger is 0.059e2.546 bar.
The pressure loss is about 20% less in the metal-foam plate heat
exchanger. In addition, the hot-side outlet temperature is
decreasing linearly with higher mass flow rate. The hot-side outlet
temperature is decreasing more slowly in the shell-and-tube heat
exchanger (Fig. 22), but the cold-side outlet temperature is
decreasing more quickly (Fig. 23).
Similarly, the heat transfer coefficient of the metal-foam plate
heat exchanger is more than 50% higher (Fig. 24). The heat transfer
Fig. 23. Cold-side outlet temperature according to cold source mass flow rate.
12
Fig. 24. The heat transfer coefficient according to cold source mass flow rate.
Fig. 25. The amount of heat transferred according to cold source mass flow rate.
rate of both heat exchangers is compared in Fig. 25. The metal-foam
plate heat exchanger has a higher heat transfer rate as the cold
source mass flow rate is increased. The rate of the metal-foam plate
heat exchanger is up to 10% higher at a mass flow rate of 250 kg/s.
5. Conclusions
A new concept for a compact heat exchanger with a metal foam
structure medium has been proposed in order to minimize the
volume and mass and to increase higher thermal efficiency in an
LNG regasification system. Numerical studies have been conducted
to design a metal-foam filled plate heat exchanger and shell-and-
tube heat exchanger using published experimental correlations.
The geometry of both heat exchangers was optimized taking into
account the conditions of thermolators in an LNG regasification
system. The heat transfer and pressure drop performance was
predicted to compare the heat exchangers. The results show that
the metal-foam plate heat exchanger has the best performance at
different channel heights according to the mass flow rate of the
fluid. In the optimized configuration of both heat exchangers, the
metal-foam plate heat exchanger has two times higher heat
transfer rate and 20% lower pressure drop than the shell-and-tube
Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056
D.Y. Kim et al. / Energy xxx (2015) 1e13
heat exchanger. The performance is increased as the mass flow rate
of natural gas is increased.
Acknowledgments
This research was supported by the National Research Founda-
tion of Korea (NRF) through Global Ph.D. Fellowship Program fun-
ded by the Ministry of Education (NRF-2014H1A2A1022186) and
through GCRC-SOP by the Korea government (MSIP) (No. 2011-
0030013). This work was also supported by the Energy Efficiency &
Resources Core Technology Program of the Korea Institute of En-
ergy Technology Evaluation and Planning (KETEP) granted financial
resource from the Ministry of Trade, Industry & Energy, Republic of
Korea (No. 20132020000390).
References
[1] Choi IH, Lee S, Seo Y, Chang D. Analysis and optimization of cascade Rankine
cycle for liquefied natural gas cold energy recovery. Energy 2013;61:179e95.
[2] Franco A, Casarosa C. Thermodynamic analysis of direct expansion configu-
rations for electricity production by LNG cold energy recovery. Appl Therm
Eng 2015;78:649e57.
[3] Gomez MR, Garcia RF, Gomez JR, Carril JC. Review of thermal cycles exploiting
the exergy of liquefied natural gas in the regasification process. Renew Sust
Energy Rev 2014;38:781e95.
[4] Gomez MR, Garcia RF, Gomez JR, Carril JC. Thermodynamic analysis of a
Brayton cycle and Rankine cycle arranged in series exploiting the cold exergy
of LNG (liquefied natural gas). Energy 2014;66:927e37.
[5] Kim KH, Kim KC. Thermodynamic performance analysis of a combined power
cycle using low grade heat source and LNG cold energy. Appl Therm Eng
2014;70(1):50e60.
[6] Koku O, Perry S, Kim JK. Techno-economic evaluation for the heat integration
of vaporisation cold energy in natural gas processing. Appl Energy 2014;114:
250e61.
[7] Liu YN, Guo KH. A novel cryogenic power cycle for LNG cold energy recovery.
Energy 2011;36(5):2828e33.
[8] Oliveti G, Arcuri N, Bruno R, De Simone M. A rational thermodynamic use of
liquefied natural gas in a waste incinerator plant. Appl Therm Eng 2012;35:
134e44.
[9] Querol E, Gonzalez-Regueral B, Garcia-Torrent J, Ramos A. Available power
generation cycles to be coupled with the liquid natural gas (LNG) vaporization
process in a Spanish LNG terminal. Appl Energy 2011;88(7):2382e90.
[10] Rao WJ, Zhao LJ, Liu C, Zhang MG. A combined cycle utilizing LNG and low-
temperature solar energy. Appl Therm Eng 2013;60(1e2):51e60.
[11] Wang H, Shi XJ, Che DF. Thermodynamic optimization of the operating pa-
rameters for a combined power cycle utilizing low-temperature waste heat
and LNG cold energy. Appl Therm Eng 2013;59(1e2):490e7.
[12] Tian J, Kim T, Lu TJ, Hodson HP, Queheillalt DT, Sypeck DJ, et al. The effects of
topology upon fluid-flow and heat-transfer within cellular copper structures.
Int J Heat Mass Transf 2004;47(14e16):3171e86.
[13] Paek JW, Kang BH, Kim SY, Hyun JM. Effective thermal conductivity and
permeability of aluminum foam materials. Int J Thermophys 2000;21(2):
453e64.
[14] Bhattacharya A, Calmidi VV, Mahajan RL. Thermophysical properties of high
porosity metal foams. Int J Heat Mass Transf 2002;45(5):1017e31.
[15] Calmidi VV, Mahajan RL. The effective thermal conductivity of high porosity
fibrous metal foams. J Heat Transf T ASME 1999;121(2):466e71.
[16] Liu JF, Wu WT, Chiu WC, Hsieh WH. Measurement and correlation of friction
characteristic of flow through foam matrixes. Exp Therm Fluid Sci 2006;30(4):
329e36.
[17] Lu W, Zhao CY, Tassou SA. Thermal analysis on metal-foam filled heat ex-
changers. Part I: metal-foam filled pipes. Int J Heat Mass Transf
2006;49(15e16):2751e61.
[18] Tadrist L, Miscevic M, Rahli O, Topin F. About the use of fibrous materials in
compact heat exchangers. Exp Therm Fluid Sci 2004;28(2e3):193e9.
[19] Mahjoob S, Vafai K. A synthesis of fluid and thermal transport models for
metal foam heat exchangers. Int J Heat Mass Transf 2008;51(15e16):
3701e11.
[20] Kim SY, Kang BH, Kim JH. Forced convection from aluminum foam materials
in an asymmetrically heated channel. Int J Heat Mass Transf 2001;44(7):
1451e4.
[21] Kim SY, Paek JW, Kang BH. Flow and heat transfer correlations for porous fin
in a plate-fin heat exchanger. J Heat Transf T ASME 2000;122(3):572e8.
[22] Iwasaki M, Asada K. Intermediate fluid type vaporizer. Google Patents; 2002.
[23] Iwasaki M, Egashira S, Oda T, Asada K, Sugino K. Intermediate fluid type
vaporizer, and natural gas supply method using the vaporizer. Google Patents;
2000.
[24] Patel D, Mak J, Rivera D, Angtuaco J. LNG vaporizer selection based on site
ambient conditions. Proc LNG 2013:17.
 D.Y. Kim et al. / Energy xxx (2015) 1e13 13

[25] Pu L, Qu ZG, Bai YH, Qi D, Song K, Yi P. Thermal performance analysis of in-
termediate fluid vaporizer for liquefied natural gas. Appl Therm Eng
2014;65(1e2):564e74.
[26] Xu SQ, Cheng Q, Zhuang LJ, Tang B, Ren QL, Zhang XJ. LNG vaporizers using
various refrigerants as intermediate fluid: comparison of the required heat
transfer area. J Nat Gas Sci Eng 2015;25:1e9.
[27] Shah RK, Sekuli c DP. Fundamentals of heat exchanger design. Hoboken, NJ:
John Wiley & Sons; 2003.
[28] Kaviany M. Principles of heat transfer in porous media. 2nd ed. New York:
Springer-Verlag; 1995.
[29] Lemmon EW, Huber ML, McLinden MO. NIST standard reference database 23.
Reference fluid thermodynamic and transport properties (REFPROP). 2012.
[30] MATLAB and statistics release. Natick, Massachusetts, United States: The
Mathworks, Inc.; 2012.

Please cite this article in press as: Kim DY, et al., Application of metal foam heat exchangers for a high-performance liquefied natural gas
regasification system, Energy (2015), http://dx.doi.org/10.1016/j.energy.2015.10.056