See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/287384766

Compact heat exchanger for latent heat recovery of exhaust flue gas

Article · January 2001

CITATIONS READS
0 367

3 authors, including:

Masahiro Osakabe
Tokyo University of Marine Science and Technology
164 PUBLICATIONS 586 CITATIONS

SEE PROFILE

All content following this page was uploaded by Masahiro Osakabe on 01 May 2017.

The user has requested enhancement of the downloaded file.

 COMPACT HEAT EXCHANGER FOR LATENT HEAT RECOVERY OF EXHAUST FLUE GAS

Jun MIHARA, Tugue ITOH, Sachiyo HORIKI and Masahiro OSAKABE

Tokyo University of Mercantile Marine, Koutou-ku, Tokyo 135-8533, Japan

Phone & FAX +81-3-5245-7404
E-Mail osakabe＠ipc.tosho-u.ac.jp

Keywords: Latent heat, Actual flue gas, Heat and mass transfer, Condensation

ABSTRACT NOMENCLATURE
The most part of energy losses in a boiler is due to the heat C: mass concentration per fluid of a unit volume
released by the exhaust flue gas to atmosphere. The released Cp: specific heat
heat consists of sensible and latent one. Recently, for a d: outer diameter of tube
biological and environmental safety, a clean fuel such as a di: inner diameter of tube
natural gas is widely used in the boiler. As the clean fuel D: mass diffusivity
includes a lot of hydrogen instead of carbon, the exhaust flue g: acceleration due to gravity
gas includes a lot of steam accompanying with the latent heat. hV: heat transfer coefficient
So the latent heat recovery from the flue gas is very important to hC: mass transfer coefficient
improve the boiler efficiency. LW: latent heat
A prediction code for the heat exchanger to recover the Nu: Nusselt number [ = h V d / λ ]
latent heat in the flue gas has been proposed. In the present Nr: total number of stages
study, the prediction code was used on the parametric study of P: pressure
the heat exchanger design for the latent heat recovery. The q: heat flux
thermal-hydraulic behavior was calculated for several kinds of R: relative humidity of combustion air
heat exchangers using finned tubes or bare tubes. The Pr: Prandtl number [ = ν / κ ]
calculation result indicated that the most compact heat Re: Reynolds number [ = ud / ν ]
exchanger was that using the bare tube of small diameter. So the
S1: spanwise pitch
compact countercurrent cross-flow heat exchanger using bare
S2: flow-directional pitch
tubes of SUS304 was designed and constructed to prove its high
ability. The outer and inner diameter of the bare tube were 10.5 Sh: Sherwood number [ = hc d / D ]
and 8.5mm, respectively. The bare tubes were arranged in a Sc: Schmidt number [ = ν / D ]
staggered bank of 10-9 rows and 40 stages. T: temperature
The experimental study varying the air ratio of flue gas, u: velocity at minimum flow area
feed water temperature and flow rate was conducted. The V: volumetric flow rate
temperature distributions of water and flue gas in the heat w: mass concentration per fluid of an unit mass
exchanger were measured. The pressure loss and the total x: relative humidity of air
amount of condensate generated in the heat exchanger were also κ: thermal diffusivity [ = λ / ( ρCP ) ]
measured. The experimental results for the temperature
λ: heat conductivity
distributions of water and flue gas in the heat exchanger with
µ: air ratio
bare tubes of small diameter agreed well with the prediction.
ν: kinematic viscosity
The proposed compact heat exchanger using small tubes was
considered to be preferable for the latent heat recovery from the ρ: density
flue gas and the prediction code was useful for the design of the
compact heat exchanger. subscript
a: atmosphere, C: condensation, COX: carbon dioxide and
monoxide, d: dry gas, F: fuel, f: flue gas, i: condensation surface,
V: convection, W: wall, N: standard condition at 0°C and
atmospheric pressure, sat: saturated condition of steam,

1 Copyright © 2001 by ASME

sub: subcooling, wt: wet gas staggered bank of 10-9 rows and 40 stages.
The experimental study varying the air ratio of flue gas,
INTRODUCTION feed water temperature and flow rate was conducted. The
The most part of energy losses in a boiler is due to the heat temperature distributions of water and flue gas in the heat
released by the exhaust flue gas to atmosphere. The released exchanger were measured with sheathed K-type thermocouples
heat consists of sensible and latent one. Recently, for a of 0.5 mm in diameter. The pressure loss and the total amount of
biological and environmental safety, a clean fuel such as a condensate generated in the heat exchanger were also measured.
natural gas is widely used in the boiler. As the clean fuel
includes a lot of hydrogen instead of carbon, the exhaust flue Table1. Composition of natural gas fuel(13A)
gas includes a lot of steam accompanied with the latent heat. So CH4 88.0 %
the latent heat recovery from the flue gas is very important to C2H6 5.8
improve the boiler efficiency. C3H8 4.5
In the previous studies [Osakabe, 1998a, 1999a, 1999b], C4H10 1.7
condensation heat transfer on horizontal stainless steel tubes has
been investigated experimentally by using the actual flue gas
CONSTITUTIVE EQUATIONS FOR PREDICTION
from a natural gas boiler. The experiments were conducted
Fuel combustion
using single and 2 stages of tubes at different air ratios and
The composition of natural gas fuel used in the test boiler is
steam mass concentrations of the flue gas in a wide range of
shown in Table.1. Volumetric concentrations of N2, CO2, O2
tube wall temperature. The condensation heat transfer was well
and CO in the dry gas were measured by a gas analyzer. By
predicted with a simple analogy correlation in the high wall
using the measured concentration, the air ratio µ is calculated as
temperature region and the low steam mass concentration typical
in air combustion. In the low wall temperature region less than N2
30°C or the high steam mass concentration typical in oxygen µ= (1)
0.79
combustion, the total heat transfer was higher than that predicted N2 − (O2 − 0.5 ⋅ CO)
by the simple analogy correlation. The condensation heat 0.21
transfer on spirally finned tubes of single stage was also The molar fraction of the carbon in the fuel of 1 mol can be
investigated experimentally and theoretically [Osakabe, 1998b]. calculated as,
The fin efficiency at the condensing region was significantly CCR = 1 × 0.88 + 2 × 0.058 + 3 × 0.045 + 4 × 0.017 mol
lower than that at the dry region. The analogous heat and mass
By using the volumetric flow rate of the fuel VF, the volumetric
transfer model for the finned tube was proposed.
flow rate of the carbon dioxide and monoxide, VCOX , is
Also in the previous studies[Osakabe, 1999c, 2000], a
prediction code for the heat exchanger to recover the latent heat V = V ⋅ CCR (2)
COX F
in the flue gas was proposed. In the prediction, the flue gas was The volumetric flow rate of dry gas Vd is,
treated as a mixture of CO2, CO, O2, N2 and H2O, and the VCOX
one-dimensional heat and mass balance calculation along the Vd = (3)
flow direction of flue gas was adapted. The heat and mass CO 2 + CO
transfer on tubes was evaluated with a simple analogy The molar fraction of the hydrogen corresponding to COX of 1
correlation. For the finned tubes, the fin efficiency at the mol can be calculated as,
condensing region was calculated with a semi-empirical CHR = ( 4 × 0.88 + 6 × 0.058 + 8 × 0.045 + 10 × 0.017 ) / CCR mol
correlation obtained in the previous basic study. The effect of
condensate film on the tubes was considered to be negligibly
Considering that the air flow rate necessary for the combustion
small for the heat transfer and pressure loss calculation in the
of the fuel is VdN2/0.79 , the volumetric fraction of H2O in the
latent heat recovery.
flue gas can be estimated as
In the present study, the prediction code was used on the
parametric study of the heat exchanger design for the latent heat (CO2 + CO) ⋅ CHR / 2 + N 2 ⋅ R ⋅ Psat /(Pa ⋅ 0.79)
H2 O = (4)
recovery. The thermal-hydraulic behavior was calculated for 1 + (CO2 + CO) ⋅ CHR / 2 + N2 ⋅ R ⋅ Psat /(Pa ⋅ 0.79)
several kinds of heat exchangers using finned tubes or bare
where Psat is the saturation pressure of steam in combustion air.
tubes. The calculation result indicated that the most compact
The volumetric flow rate of wet gas, Vwt, is
heat exchanger was that using the bare tube of small diameter.
So the compact countercurrent cross-flow heat exchanger using Vd
Vwt = (5)
bare tubes of SUS304 was designed and constructed to prove its 1 − H2O
high ability. The outer and inner diameter of the bare tube were
10.5 and 8.5mm, respectively. The bare tubes were arranged in a When the flue gas temperature is Tf°C, the steam mass

2 Copyright © 2001 by ASME

concentration Cf per unit volume of flue gas is,  κ 
D = D air   (15)
κ 
Cf =
H2 O ⋅ 18
⋅
273.15
(6)
 air 
22.4 273.15 + Tf where κ and κair are the thermal diffusivities of flue gas and air,
respectively. The diffusivity of steam in air can be expressed as
The steam mass concentration Wf per unit mass of flue gas is
(Fujii et. al., 1977),
H2 O ⋅ 18
wf = 11/ 6
H2 O ⋅ 18 + (1 − H2 O )( CO 2 ⋅ 44 + CO ⋅ 28 + N2 ⋅ 28 + O 2 ⋅ 32 ) −5 ( T + 273.15 )
D air = 7.65 × 10 (16)
(7) P
The previous experimental study for single stage of finned
Heat and mass transfer in gas side tube [Osakabe, 1998b] showed the following empirical
The total heat flux qT consists of the convection heat flux correlation was available in the range of 2×103＜Ref≦5×105.
qV and the condensation heat flux qC as 0.33
Nu f = j Re f Prf (17)
q T = q V + qC (8)
0.5
The convection heat flux is expressed as  d + LF 
where j = C1C 3 C 5   (18)
q V = h V ( Tf − Tw ) (9)  d 
The condensation heat flux can be expressed as, −0.35
C1 = 0.25 Re f (19)
q C = h CL W ( C f − C w ) (10)
−0.25L / S
where Cw is the mass concentration of saturated steam at the wall C 3 = 0.35 + 0.65e F F (20)
temperature Tw. Based on the previous studies, the Nusselt
number Nuf for the average convective heat transfer coefficient C 5 = 0.7 (21)
in the range of 103＜Ref≦2×105 is The fin efficiency can be calculated by,
0.6 m 0.25   d + LF  
Nu f = c Re f Prf (Prf / Prw ) (11)
η = YF 0.45 ln ( YF − 1) + 1 (22)
Zukauskas(1972) proposed m=0.36 and   d  
0.2
For S1/S2＜2 c = 0.35(S1 / S 2 ) (12) where YF = XF (0.7 + 0.3 XF ) (23)

For S1/S2≧2 c = 0.40 (13) tanh(mb )

XF = (24)
mb
for a staggered bank. For an analogous mass transfer process, 0.5
the Nusselt number and Prandtl number in the heat transfer  2h 
relation Eq.(11) are simply replaced by the Sherwood number m=  (25)
and the Schmidt number, respectively. This procedure gives  λF t F 
0.6 m 0.25 b = LF + t F / 2 (26)
Sh f = c Re f Sc f (Sc f / Sc w ) (14)
The rule gives a correct relation for the limiting situation in The fin efficiency strongly depends on the heat transfer
which the differences in temperature and concentration are coefficient, h, in Eq.(25). These correlations were obtained from
vanishingly small, and it is valid for independent analogous heat the single phase experiment where the heat flux is the multiple
and mass transfer situations as well as for a combined heat and of the constant heat transfer coefficient and the temperature
mass transfer process. difference between the wall and the fluid. Although the
Flue gas was treated as a mixture of N2, CO2, O2, CO and condensation heat flux can not be expressed with the above
H2O and its property was estimated with special combinations of simple relation, it is considered that the Eq.(25) with the larger
each gas property proposed by the previous studies(JSME, heat transfer coefficient taking account of the condensation heat
1983). For example, the heat conductivity and the viscosity were transfer would give an approximation. The previous study
estimated with the methods by Lindsay&Bromley(1950), and showed the fin efficiency could be evaluated with following
Wilke(1950), respectively. It is considered that a strong equivalent heat transfer coefficient in the condensation region.
correlation exists between the thermal and mass diffusivities. As h L W (C f − C W )
a first attempt, the mass diffusivity of steam in flue gas was h = hV + β C (27)
estimated with the well-known mass diffusivity of steam in air as Tf − TW
where β=1 for the usual calculation.

3 Copyright © 2001 by ASME

 For an analogous mass transfer, the simple analogy Heat transfer in water side
between heat and mass transfer gives, Heat transfer correlation by Dittus-Boelter taking account
0.33 of the pipe inlet region is used. The inlet coefficient by
Sh f = j Re f Sc f (28)
McAdams(1954) is used for the modification.
The one-dimensional heat and mass balance calculation
along the flow direction of flue gas was conducted. The steam 0.8 0.4  di 0.7 

Nu = 0.023 Re Pr  1 + ( ) (35)
mass concentration and the flue gas temperature at N+1th stage  L 
can be calculated from those at Nth stage as; where L is the heating length of tube.
q A W 22.4
H2 O(N) ⋅ V − C ⋅ Pressure loss calculation
wt LW 18
H2 O(N + 1) = (29) The pressure loss per a stage of tube is,
qC A W 22.4 2
V − ⋅ ∆P = 2f ρ f u (36)
wt LW 18 For the staggered bank of bare tube, Jacob(1938) proposed the
qV A W following coefficient f,
Tf (N + 1) = Tf (N) −  
CPf ρ f ( 273.15 + Tf ) / 273.15 ⋅ V  
wt
(30)  0.118  −0.16
f = 0.25 +
1.08 
where Aw is the heat transfer area per a stage. Re f (37)
It is possible that the gas temperature coincides with the   S1   
dew point which is the saturation temperature corresponding to    − 1 
the partial pressure of steam in the flue gas. When the gas    d   
temperature decreases below the dew point, the condensation of For the staggered bank of finned tube, the empirical correlation
steam in the flue gas takes place and the latent heat increases the by ESCOA(1979) is,
gas temperature until the gas temperature coincides with the dew 0.5
point. In this case, the energy balance gives the relation between  d + LF 
f = C 2 C 4 C 6   (38)
the increase of the gas temperature, ∆Tf, and the decrease of  d 
steam concentration, ∆H2O, as; −0.35
18 LW where C 2 = 0.07 + 8 Re f (39)
∆Tf = ⋅ ⋅ ∆H2 O (31)
22.4 CPf ρ f ( 273.15 + Tf ) / 273.15 −0.7(L / S )0.20
 S  F F
The average film thickness of condensate was calculated by C 4 = 0.110.05 1  (40)
the method shown in APPENDIX. As the existence of the film  d 
did not affect significantly the prediction for the present  −0.15Nr2  −2.0S2 / S1
experiment, the thermal resistance of the film was neglected. C 6 = 1.1 + 1.8 − 2.1e e
  (41)
Heat conduction in tube  −0.15Nr2  −0.6S2 / S1
The heat conductivity for the inconel or austenite stainless − 0.7 − 0.8e e
steel is given with the following approximate correlation  
[Osakabe, 1980].
λ t = 13.2 + 0.013Tt W/(m K) (32) PARAMETRIC STUDY
Heat transfer tubes were installed in the rectangular duct of
where Tt is the average temperature of tube as, 205×205mm to recover the latent heat in flue gas as shown in
T + TWi Fig.1. This duct size is approximately the same as the flue gas
Tt = W (33) duct of test boiler used in this experiment. The outlet
2
temperature of flue gas from the boiler was 280℃. The flue gas
where Tw and Twi are the outer and inner wall temperatures,
respectively. The heat flux at the outer wall is, was generated with natural gas 13A at the flow rate of 15 mN3/h
and the air ratio of 1.2. The flow rate and inlet temperature of
2λ t ( Tw − Twi )
qw = (34) feed water was 600kg/h and 20℃, respectively. The temperature
d ln( d / di ) of the feed water was increased from 20℃ to 60℃ with the
heat recovery.
The various kinds of heat transfer tubes are installed in the

4 Copyright © 2001 by ASME

duct as the parametric design study. Bare tubes of 21.7, 10.5 and Table2. Comparison of bare and finned tubes
5mm in outer diameter, spirally finned tubes are selected in the Tube type Bare1 Bare2 Fin1 Fin2 Fin3
study. The outer diameter of finned tubes is 21.7 mm and the fin Pipe diameter(mm) 10.5 21.7 21.7 21.7 21.7
heights are 12, 8 and 3mm. The pitch of tube arrangement is Pitch (mm) 20.5 34.2 51.0 41.0 34.2
larger than the tube outer diameter by 10mm considering the Stage 35 40 29 25 27
fabrication ability of holes at the tube sheets. The maximum Pipe No. 333 220 102 113 149
number of tubes is 333 using the tube of 10.5 mm in outer Fin Pitch 5 5 5
diameter. The minimum number is 102 using the finned tube of Fin Height 12 8 3
12mm in height, which is often used in conventional economizer
L (mm) 718 1370 1480 1030 923
for the sensible heat recovery. The most compact economizer is
A(m2) 2.25 3.07 12.4 8.73 5.04
that using small bare tubes of 10.5mm in diameter. The height
W(kg) 18.9 46.1 62.6 51.1 42.5
and the total weight of the economizer using small bare tubes
Gas Pressure loss 11.5 21.6 19.8 22.1 22.4
are 718mm and 18.9kg, respectively. The pressure loss in the
(mmAq)
water side is slightly higher than the others but can be allowable.
On the other hand, the total weight of that using the finned tubes Water Pressure loss 515.0 65.3 117.0 60.5 43.9
of 12mm in height was 62.6kg. The bare tube of small diameter (mmAq)
is preferable for the compact design of economizer for the latent
heat recovery. Table.3 Experimental conditions
Air ratio 1.41 1.4 1.37 1.46
205 205 Feed water (kg/h) 610 537 482 604
13A Fuel (mN3/h) 16.1 15.86 15.9 16.07
13A Fuel Gas Gas inlet (℃) 287 289 292 290
Gas inlet Water (℃)
15mN3/h 21→67 18→71 19→76 16→67
280℃ Gas Pressure loss (mmAq) 19 19 19 23
Air ratio=1.2
L
Feed water Water Gas
600kg/h 20→60℃ Water
Header
Tubes: SUS304, Sch10S

Fig. 1 Boundary conditions for parametric study

1476
COMPACT HEAT EXCANGER EXPERIMENT
40
Shown Fig.4 and 5 are a schematic of experimental stages Water
apparatus. The multiple stage experiment was conducted by
using a flue gas generated with the combustion of air and natural
gas fuel. The flue gas from a natural gas boiler is led to the inlet
plenum of the test heat exchanger using small tubes. The flue
gas was released to atmosphere from the outlet plenum. The 260
countercurrent cross-flow heat exchanger, which consist of bare
tubes outer and inner diameter 10.5 and 8.5mm respectively,
were designed and used for the experiment. The effective 220
heating length of the bare tubes was 200mm.
And the small bare tubes were arranged as a bank of 10-9
Fig. 2 Schematic of experimental apparatus for
rows and 40 stages. The temperature distributions of water and
compact heat exchanger
flue gas in the heat exchanger were measured with sheath T-type
thermocouples of 0.5mm in diameter. The thermocouple signals
were transferred to a date logger and analyzed. The
measurement error of the temperature in this study was within
±0.1K. The pressure loss and the total amount of condensate
generated in the heat exchanger were also measured. The major
experimental conditions are shown in Table.3.

5 Copyright © 2001 by ASME

 exchanger. The measured total amount of condensate is slightly
higher than the prediction with the one-dimensional mass and
Water inlet heat balance calculation.
880
950 Thermocouple 830
30

25 Air ratio 1.41

Air ratio 1.49

Predicted Δp(mmAq)
20
200 Water outlet 220
220 15

Fig. 3 Cross-section of compact heat exchanger using 10

Small Tubes
5
3
0
0

0
GOI
a
su
erwrr
t
mwat
p
al
r
a
tmmrmsa
u
r
ep

0 5 10 15 20 25 30
t
ee
al
ll
t
e
aa
r
a
tu uam
u
r
ee ea
2
5
0

Measured Δp(mmAq)
n
na
rt

t
e
pa
rr
a
t
r
WS
eu
edde
mt d
p
a
t
ua
ea
a
te
tu
a
e
eaw
pt
re
t
r
2
0
0

Fig. 5 Pressure difference between inlet and outlet of

MM
sa
a
ru
e
g

mt
p
r
t
u
r
et
Temperature(℃)

e
s
r

t
e
r
e
p
a
r
a
u
r
e

heat exchanger
A
i
re
r
a
t
i
oa
1
.
4
96
1
5
0

F
e
d
w
t
e
r
0
0
k
g
/
h

20
1
0
0

18
Predicted condensate (kg/h) 16
5
0

14
0

12
0

1
0

2
0g

3
0

4
0
S
t
a
e
s

10
8
6
Water flow rate 10 kg/min.
Fig. 4 Temperature distribution in compact heat 4 Water flow rate 9 kg/min.
exchanger 2 Water flow rate 8 kg/min.

0
Shown Fig.4.is the comparison of experimental results and 0 2 4 6 8 10 12 14 16 18 20
Measured condensate (kg/h)
predictions. The solid lines are the prediction of gas and water
temperature. The dashed line is the saturation temperature
Fig. 6 Predicted and measured amount of
corresponding to the partial pressure of steam in the flue gas.
condensate
The gas temperature merges and decreases with the saturation
temperature. The measured temperature of water (□) and flue
CONCLUSION
gas ( ● ) agree well with the predictions. The effect of
condensate film on the tubes was considered to be negligibly 1. The prediction code was used on the parametric study of the
small for the heat transfer calculation. heat exchanger design for the latent heat recovery. The
Shown in Fig.5 is the comparison of experimental result thermal-hydraulic behavior was calculated for several kinds
and prediction for the pressure loss throughout the heat of heat exchangers using finned tubes or bare tubes. The
exchanger. As the empirical correlation obtained in the calculation result indicated that the most compact heat
non-condensing region was used in the one-dimensional exchanger was that using the bare tube of small diameter.
calculation, the experimental results for the banks of small tubes 2. The compact countercurrent cross-flow heat exchanger
is a slightly higher than the prediction. But the effect of using bare tubes of SUS304 was designed and constructed
condensate film on the tubes is considered to be negligibly small to prove its high ability. The outer and inner diameter of the
for the pressure loss calculation. bare tube were 10.5 and 8.5mm, respectively. The bare
Shown in Fig.6 is the comparison of experimental result tubes were arranged in a staggered bank of 10-9 rows and
and prediction for the amount of condensate throughout the heat

6 Copyright © 2001 by ASME

 40 stages. 286-294.
3. The experimental study varying the air ratio of flue gas, Wilke, C.R., 1950, “A viscosity equation for gas mixture”,
feed water temperature and flow rate was conducted. The J. Chem. Phys., 18, 517-519.
experimental results for the temperature distributions of Zukauskas, A., 1972, Advances in Heat Transfer, 8,
water and flue gas in the heat exchanger with bare tubes of Academic press, New York, 93-160.
small diameter agreed well with the prediction.
4. The proposed compact heat exchanger using small tubes APPENDIX
was considered to be preferable for the latent heat recovery It is assumed that all the condensate generated at the upper
from the flue gas. stage flows on the tubes as a laminar film as shown in Fig.7. The
momentum balance dominated by viscous and gravity force
ACKNOWLEDGMENT gives the velocity distribution at θ° from the tube top:
The authors appreciate the helpful supports by Kawasaki
u=
( ρ L − ρ G )g sin θ  yδ − y 2 
Thermal Engineering Co. Ltd.   (42)
µL  2
REFERENCES Integrating the above velocity profile and using the condensate
ESCOA FINTUBE CORPORATION, 1979, SOLIDFIN mass flow rate per unit of tube length, m, yields
HF. 1/ 3
Fujii, T., Kato, Y and Mihara, K., 1977, “Expressions of  . µ Lm
15 
δ=  (43)
 ρ L ( ρ L − ρ G ) g sin θ 
transport and thermodynamic properties of air, steam and water”,
Univ. Kyushu Research Institute of Industrial Science Rep.66,
81-95. The heat conductivity of film is
Jakob, M., 1938, “Heat transfer and flow resistance in cross 1/ 3
λ L  λ L ρ L ( ρ L − ρ G ) g sin θ 
3
flow of gases over tube banks”, Trans. ASME, 60, 384. K= =  (44)
JSME, 1983, Data Book: Heat Transfer 3rd Edition, (in δ  . µ Lm
15 
Japanese).
Lindsay, A.L. and Bromley L.A., 1950, “Thermal The average conductivity from θ= 0° to θ=π is
1/ 3
conductivity of gas mixtures”, Indust. Engng. Chem., 42,
1
π
 λ L 3 ρ L ( ρ L − ρ G )g 
1508-1510. K = ∫ K dθ = 0.72  (45)
McAdams, W.H., 1954, Heat transmission, π 0  µ Lm 
McGRAW-HILL. The average heat resistance of film is defined as the inverse of
Osakabe, M., Ishida, K., Yagi, K., Itoh, T. and Ohmasa, M., the above average conductivity. The average film thickness is
1998a, “Condensation heat transfer on tubes in actual flue gas
λL
(Experiment using flue gas at different air ratios)”, (in Japanese), δ= (46)
Trans. of JSME , 64-626, B, 3378-3383. K
Osakabe, M., Itoh and T. Ohmasa, 1998b, “Condensation In the calculation, the mass flow rate, m, at a certain stage
heat transfer on tubes in actual flue gas (Experiment using includes the condensate generated at the stage for the
spirally finned tubes)”, (in Japanese), Proc. of MESJ 61st conservative estimation.
Meeting, 70-73.
Osakabe, M., Yagi, K., Itoh, T. and Ohmasa, M., 1999a,
“Condensation heat transfer on tubes in actual flue gas Condensate
(Parametric study for condensation behavior)”, (in Japanese),
Trans. of JSME , 65-632, B, 1409-1416. y
Osakabe, M., Itoh, T. and Yagi, K., 1999b, “Condensation
heat transfer of actual flue gas on horizontal tubes”, Proc. of 5th θ δ
ASME/JSME Joint Thermal Eng. Conf., AJTE99-6397. Tube
Osakabe, M., 1999c, “Thermal-hydraulic behavior and
prediction of heat exchanger for latent heat recovery of exhaust g
flue gas”, Proc. of ASME, HTD-Vol.364-2, 43-50
Osakabe, M., 2000, “Latent Heat Recovery from
Oxygen-Combustion Flue Gas”, Proc. of 35th Intersociety
Energy Conversion Conference, Vol.2, 804-812 Fig. 7 Heat conductance of condensate film
Osakabe, M., 1989, “Thermal-hydraulic study of integrated
steam generator in PWR”, J. Nucl. Sci. & Technol., 26(2),

7 Copyright © 2001 by ASME

View publication stats