Lehigh University

Lehigh Preserve
Theses and Dissertations

1991

Calculation of the metal temperature distribution

within a two- pass counter-crossflow tubular air
preheater
Ali Yilmaz
Lehigh University

Follow this and additional works at: http://preserve.lehigh.edu/etd

Recommended Citation
Yilmaz, Ali, "Calculation of the metal temperature distribution within a two- pass counter-crossflow tubular air preheater" (1991).
Theses and Dissertations. Paper 6.

This Thesis is brought to you for free and open access by Lehigh Preserve. It has been accepted for inclusion in Theses and Dissertations by an
authorized administrator of Lehigh Preserve. For more information, please contact preserve@lehigh.edu.
AUTHOR:
Yilmaz, Ali

TITLE:Calculation of
the Matal Temperatu"re
Distribution Within a
Two-Pass Counter-
Crossflow Tubular
Air Preheater
DATE:January 1992
CALCULATION OF THE METAL TEMPERATURE DISTRIBUTION

WITHIN A TWO-PASS COUNTER-CROSSFLOW

TUBULAR AIR PREHEATER

by

Ali Yilmaz

A Thesis

Presented to the Graduate Committee

of Lehigh University

111 Candidacy for the Degree of

1\1 aster of Science

111

i->Iechanical Engineering

Lehigh liniversity

Decem ber. 1991

 ACKN OWLEDG EM ENTS

I am very grateful to my parents and fiancee for their love, support and

understanding during my study at Lehigh. I would also like to thank my advisor,

Dr. Edward K. Levy, for his guidance and support throughout my work.

The work in this thesis was funded by Allegheny Power Company as part

of a project. I would like to express my appreciation to the sponsors for their

support.

am also grateful t.o Dr. Nenad Sarunac for his help and encouragement

during my research.

Finally. I would like to thank my friends Ahmet Pinarbasi. l\lurat Ozturk,

Patrick Coston, Ridvan A. Sahan, and Hasan Gunes for their support and help.

III
 TABLE OF CONTENTS

Page

TITLE PAGE
II
CERTIFICATE OF APPROVAL

ACKNOWLEDGEMENTS iii

IV
TABLE OF CONTENTS
VII
LIST OF TABLES

LIST OF FIGURES viii

XI
NOMENCLATURE

ABSTRACT 1

INTRODUCTION 2
1.

2. HEAT TRANSFER IN DIRECT-TYPE HEAT EXCHANGERS 5

2.1 Dimensional and Nondimensional Variables 5

2.2 Assumptions Applied to (-NTU Theory 10

2.3 Longitudinal Heat Cond uction Effect on Air

Preheater Performance 10

CONVECTION CORRELATIONS 18
3.

3.1 Introduction 18

3.2 Flow Inside Circular Tu be 18

3.2.1 Laminar Flow 19

3.2.2 Turbulent Flo,,' 20

3.2.3 The Influence of Surface Roughness

3.3 Flow Over Tube Banks 29

IV
 3.3.1 Banks of Tubes 29

3.3.2 Mean Heat Transfer from Banks of Tubes 30

3.3.3 The Influence of Roughness on the Heat Transfer 34

4. MATHEMATICAL MODEL 37

4.1 Single-Pass Crossflow Exchanger 37

4.2 Verification of the Model 41

4.3 Two-Pass Counter-Crossflow Air Preheater 42

5. EFFECT OF AIR LEAKAGE ON AIR

PREHEATERPERFORMANCE

5.1 Air Leakage Model

5.2 Air Preheater Performance with Leakage 59

6. COMPUTER ALGORITHM 68

6.1 Introduction 68

6.2 General Approach in the TPHMT Code 69

6.2.1 Input Data

6.2.2 Possible Applications of the Computer Program

6.3 Results of t he Sample Calculations 73

6.3.1 Effect of Inlet Air Temperature 73

6.3.2 Effect of Inlet Gas Temperature 74

6.3.3 Effect of Gas Flo\\" Rate 74

6.3.2 Effect of Air Flo\\" Rate 74

I. SU ~ I T\ 1A R\' A)\' D CO j\; eLl: S I0 j\; S 88

8. RECOMMENDATIONS FOR FUTURE WORK 89

9. REFERENCES 92

APPENDIX A 95

VITA 107
 LIST OF TABLES

Page

Table 3.1 Results of the heat transfer analysis for simultaneously

developing laminar flow inside a circular tube (Pr=0.7) [13]. 20

Table 3.2 Ratio of mean to fully developed turbulent flow Nusselt

number in the entrance region of a smooth circular tube

with various entrance configurations for Pr=0.7 [18]. 24

Table 3.3 Equivalent sand roughness for commercial circular tubes [21]. 26

Table 3.4 Constants of equation 3.14 for tube banks in crossflow [23]. 32

Table 4.1 Comparison between the effectiveness of the mathematical

model presented in Chapter 4.1 and the analytical solution [3] for

a single pass crossflow exchanger having both fluids unmixed. 47

Table 5.1 Results of air leakage analysis. 62

VII
 LIST OF FIGURES

Page

Figure 2.1 Effectiveness of a single-pass crossflow heat exchanger with

one of the fluids mixed and the other unmixea [20]. 9

Figure 2.2 The dimensions of one tube of a crossflow tubular heat exchanger.

The longitudinal heat conduction effect was studied using this

particular tube (Chapter 2.3). 12

Figure 2.3 J\'ondimensional temperature profiles along the one tu be of the

crossflow exchanger; the t.wo wit.h longit.udinal heat conduction

(Ag=0.00031),eqns 2.26 and 2.27, the other two without (.\g=O),

eqns 2.34 and 2.35. 17

Figure 3.1 Friction factor for fully developed flow in a circula,r tube [21]. 27

Figure 3.2 Tube arrangement.s in a tube bank; a) in-line, b) staggered. 30

Figure 3.3 Flow conditions in a tube bank; a) in-line, b) staggered. 36

Figure 4.1 Subdivisions of a single-pass crossflow heat. exchanger on xy plane. 38

Figure 4.2 The effect of grid sIze on t.he heat. exchanger effectiveness. 43

Figure 4.3 The effect of grid size on the accuracy of the mat.hematical model

results for heat exchanger effectiveness. 44

Figure 4.4 The effect of grid sIze on the total heat transfer rate. 45

Figure 4.5 The effect of grid sIze on t he accuracy of the mathematical model

results for heat transfer rate of the heat exchanger. ·16

Figure 4.6 The sketch of the tubular recuperative air preheater arranged as a

VI1I
 two-pass counter-crossflow. 49

Figure 4.7 Subdivisions of Pass I of the air preheater on xy plane. 51

Figure 5.1 Internal and external mass flow rates. 56

Figure 5.2 Internal and external fluid temperatures. 57

Figure 5.3 The air leakage effect on the air preheater performance. 61

Figure 5.4 The air leakage effect on outlet gas and air temperatures of the air

preheater. 62

Figure 5.5 The air leakage effect on mass flow rates of the air preheater. 63

Figure 6.1 Flow chart of computer code TPHMT. 66

Figure 6.2 The sketch of the tubular air preheater. 67

Figure 6.3 The effect of the inlet gas and air temperatures on the

overall effectiveness of the air preheater. 76

Figure 6.4 The effect of the inlet gas and air temperatures on the outlet all'

temperature of the air preheater. 77

Figure 6.5 The effect of the inlet gas and air temperatures on the outlet gas

temperature of the air preheater. 78

Figure 6.6 t\letal temperature variation with the inlet gas and air temperatures

of the air preheater at x=.24 m and y=9.21 m within Pass 1. 79

Figure 6.7 Metal temperature variation with the inlet gas and air temperatures

of the air preheater at x=.31 m and y=.48 711 within Pass If. 80

Figure 6.8 The effects of the mass flow rat.es of gas and air on the outlet gas

temperature of t.he air preheat.er. 81

Figure 6.9 The effects of the mass flow rates of gas and aIr on the outlet aIr

temperature of the air preheater.

IX
Figure 6.10 The effects of the inlet gas and air mass flow rates on the

overall effectiveness of the air preheater. 83

Figure 6.11 Metal temperature variation with the inlet gas and air mass flow

rates of the air preheater at =.24 m and y=9.21 m

within Pass I. 84

Figure 6.12 Metal temperature variation with the inlet gas and air mass flow

rates of the air preheater at =.31 m and y=.48 m

within Pass II. 85

Figure 6.13 J\letal temperature distribution within Pass I

for baseline conditions. 86

Figure 6.14 J\letal temperature distribution within Pass II

for baseline conditions. 87

Figure 8.1 i\lulti-pass crossflow air preheaters; a) three-pass counter-crossflo\\',

b) four-pass cOllnter-crossOo\\'. 90

Figure 8.2 a) Possible flow directions. and b) possible givcn boundary

condit.ions of a single-pass crossOow cxchanger. 91

 NOMENCLATURE

a The relative transverse pitch, STI D

')
A Exchanger total heat transfer area on one side, m-

->,--...
Total solid area for longitudinal conduction along a tube,
') ') ')
Aw=iT(ra--rg-), m-

b The relative longitudinal pitch, S LI D

b1 , b'2' b3 Coefficients in Equation '2.'27

C The relative diagonal pitch, S DI D

c 1' c'2' c3 Coefficients in Equation '2.'26

cp Specific heat at constant pressure, J Ikg K

C Constant in Equation 3.6; Constant in Equation 3.14

Ca , Cg Capacity rates of air and gas streams, VV IK

C min , C max Minumum and maximum capacity rates, \\1 II<

CR Capacity rate ratio. C R = Cminl C max

D Diameter of a circular tube, m

D Hydraulic diameter of any internal passage, m

h

I Friction factor, I=7 w l(§Pu m'2)

Is Friction factor for smooth tube surfaces

h Convectivc heat transfer coefficient. \V/m'2 K

'- Thcrmal conductivity of working fluids. \V 1m I":

'-w Thermal conductivity of thc tube mctal. \\'/m K

L Tube lcngth. m

Xl
 Mass flow rates of air and gas, respectively, kgjs

m Constant in Equation 3.14

e External mass flow rate of air at the inlet of Pass I, kgjs

mail

Internal mass flow rate of air in Pass I, kg/s

mapass Passage mass flow rate of air, kgjs

1
m a2 Internal mass flow rate of air in Pass II, kgjs

Mass flow rate of air at the outlet of Pass II, kgjs

External mass flow rate of gas at the inlet of Pass II, kg/s

1
m ') Internal mass flow rate of gas in Pass II, kgjs
g-

mgpass Passage mass flow rat.e of ga.s, kg/s

1
m gl Internal mass flow rat.e of gas in Pass J, kg/s

e External mass flow rat.e of gas at. t.he outlet of Pass I. kg/s
m gol

m L.eei Air leakage mass flow rate through gas-in section in t.he cold end of

either pass, kgjs

771 L.eeo Air leakage mass flow rat.e through gas-out section In the cold end of

ei t her pass. kg/ s

771 L.hei Air leakage mass flow rat.e through ga.s-in section 111 t he hot end of

eit.her pass. kg/s

771 L.heo Air leakage mass flow rate through gas-out sect.ion In the hot end of

eit her pass. kg/s

m L . CE Cold end air leakage mass flow rate on either pass. kg/s

mL.HE Hot. end air leakage mass flow rate on either pass. kg/s

m L1 · 171 Air leakage mass flow rates of Pass J and Pass II. respecti\·ely. kg/s
L2

mL Total air leakage flow rate of the air preheater. kg/s

XII
M Index

N Index

n Constant in Equation 3.6

NTU Number of heat transfer units, Eqn. 2.6

Nu Nusselt number

NUfd Nusselt number for fully developed flow and heat transfer

NUm Mean Nusselt number

Nus Nusselt number for smooth surfaces

Nux Local Nusselt number

NU m 1-2 Average Nusselt number between any arbitrary two points

Pr Prandtl number

Prs Prandtl number based on the tube wall temperature

qa, qg Heat transfer rates on air side and gas side, respectively, \V

q"w " - dq
H eat fl ux,qw-dA' \<'1
\ m2

Q Total heat transfer rat.e of the air preheater, \V

r a , rg Outside and Inside radius of a circular tube, m

ra+r g
rm Average tube radius as 7'm=~, m

The roots of the Equation 2.29

Re Reynolds number

Critical Reynolds number

pVD
ReD Reynolds number. ReD=-'-j-

Re Dmax Reynolds number based on the maxImum flow velocity in a tubc

- p V m o.r D
b an k". R e Dmox- P
711
R
· paramctcr.C - L .ceo
L ea k'age ratIo
·E- 77lL.cei

XIII
 Leakage ratio parameter, R _ m L,heo
HE-m L , h el.

fJ
Leakage ratio parameter, R c - L,GE
o-fJ -
,
LHE
S D' S L' S T Diagonal, longitudinal, and transverse pitch of a tube bank, m

T Temperature, K

Tail Air inlet temperature of Pass I, K

Tapass Passage air temperature, K

T ao2 Air outlet tempel'ature of Pass II, K

External gas temperature at the inlet of Pass II, K

Internal gas temperature at the inlet of Pass II, K

T gpass Passage gas temperature, K

1
T gz'1 Internal gas temperature at the inlet of Pass I, K

T 1 gal Internal gas temperature at the outlet of Pass I, K

C
T gal External gas temperature at the outlet of Pass I, K

u Overall heat transfer coefficient, \V 1m 2 K

Urn \lean flow velocity in a circular tube, m/s

V max \laximum flow velocity in a tube bank, mls
x, y. ;; Cartesian coordinates, III

x Thc critcria dcfincd in Eqn. 4.12

Entry Icngth of a simultancously dcveloping flow in a tube. Il1

x* D ·IlllenSlOn . a tu b e, x * = (xl
. I ess d'Istancc In RcPr
D )
h
;; Nondimensional coordinate, ;;=ZI L
Z Coordinate of a tube in the gas flO\\· direction, III

XIV
 Number of heat transfer units on gas side, o:=~L;
9
Thermal diffusivity, m 2 Is

(3 Number of heat transfer units on air side, (3= ~L

9
o£ Total air leakage coefficient of the air preheater

°£1' ° £2
Leakage coefficients of Pass I and Pass II, respectively

o£,CE Cold end leakage coefficient

o£ ,HE Hot end leakage coefficient

( Exchanger effectiveness

Heat transfer effectiveness of the tubular air preheater based on the

rate of heat transfer to air and from gas, respectively

Dimensionless temperature of air, gas, and tube metal. respectively

Longitudinal cond uction parameter defined by Eqn. 2.17

Viscosi ty, kgl s m

')
v Kinematic viscosity, m- Is

p Mass density, kg/m 3

')
'\\' Shear stress, N Im-
 ABSTRACT

A large number of tubular aIr preheaters in coal fired power plants have

been experiencing difficulty with plugging and corrosion. \""hen a flue gas

containing sulfur trioxide and water vapor contacts a cold heat transfer surface in

the air preheater, sulfuric acid will precipitate and fly ash will impact the liquid

forming a sticky, adherent deposit. If the fly ash is not sufficiently alkaline to

neutralize the precipitated sulfuric acid, in the long term the tubes will suffer from

corrosion. In order to protect the tubes from acid deposition and corrosion, we

need to keep the metal temperature field of the aIr preheater above the acid

dewpoint temperature. This requIres that the metal temperature field of the aIr

preheater be known as a function of operating parameters of the all' preheater.

Since measurements of metal temperatures are not practical except for research

purposes. a numerical simulation of the heat transfer problem IS the only

alternative.

A computer program namcd TPH1'.IT (Tubular Air Preheater :\Ietal

Temperature) lias l)('cn devcloped for a t\\"o-pass counter-crossllo\\" tuhular air

preheatcr for computing the metal temperature field within the air preheater as a

function of design and operating parameters. The mathematical mode!. based on

the (-0:Tl' theory. uscs an iterative procedure to ohtain the solutions.

 1. INTRODUCTION

Recovering waste heat from a high temperature gas stream by using an aIr

preheater to preheat incoming combustion air IS an effective way of reducing fuel

consumption. In industries such as electric power generators, steel and glass

manufacturing, air preheaters have been used for years.

Air preheaters may be classified according to the principle of their

operation as (a) recuperative and (b) regenerative. In a recuperative design, the

heat is transfered directly from the hot gases or steam, on one side of the surface,

to air on the other side; whereas III a regenerative air preheater, the heat is

transfered inderectly from the hot gases to the aIr through some intermediate

storage medium. Air preheaters operating on the recuperative principle are

generally tubular type, although some are plate type.

Experiences indicate that corrosion may occur III all' preheaters operating

with flue gas when the metal temperature of the tube falls below the sulfuric acid

dewpoint temperature. This temperature is the highest temperature at which

condensation of sulfuric acid can exist on a cooled surface in equilibrium with a

flue gas conta.ining 503 and H 2 0. Acidic condensate deposited in sections of the

flue gas passages operating at low temperature could cause rapid corrosion of the

tube materials and blockage of the gas passages with deposits of fly ash. These

problems can be a"oided if the cold end of the air preheater is allowed to operate

a few degrees hotter. This is commonly done by several means: (1) heating the

air entering the preheater, (2) recirculating a portion of the hot air from the air
preheater outlet back to the forced draft fan inlet.

Information on the metal temperature field must be known as a function

of operating parameters of the air preheater in order to avoid acid deposition and
I
corrosion problems. This information on metal temperatures can be obtained b'y

either taking direct measurements or by numerically simulating the heat transfer

processes within the air preheater. Measurements of metal temperatures are

difficult to make in a large heat exchanger; therefore, except for research purposes

it is not practical to rely on such measurements as the source of information. The

alternative way is to have a computer program which performs a numerical

simulation of the heat transfer processes within the air preheater.

In this thesis, a tubular air preheater arranged in a two-pass counter-

crossnow configuration was modelled and a computer program was written. The

numerical simulation of the heat transfer problem was based on the (-NTU

method [1].

The air preheater is su bdivided into two separate passes, each of which can

be viewed as a single pass crossflow heat exchanger (see Figure 6.2). The gas flow

between these passes was assumed to behave as a one-dimensional adiabatic flow.

That is, the temperature of the gas is assumed t.o remain unchanged t.hroughout

t.he passage which connects bot.h passes.

;-'loreover. to account for temperature and velocity nonuniformities of gas

and air at the inlet. of t.he air preheater. each pass is divided into .\lxN separate

smaller element.s (grids). Since inlet ga.s temperature and velocity variations for

one pa.ss and inlet air temperature and velocity \·ariat.ions for t.he other pa.ss were

considered as boundary conditions: an iterative prediction-correction method wa.s

3
applied to find the metal temperatures. Outlet gas and air temperatures and tube

metal temperature of each grid are directly computed by using the (-NTU

relations. The temperature fields of air, gas and tube metal are calculated using a

step bY::5tep marching method starting from an appropriate smaller element in the

first pass.

The assumption of negligible longitudinal heat conduction along the

lengths of the tubes within the air preheater made it possible to use (-NTU

method. This was justified by a simple analysis which showed that longitudinal

conduction along a tube in crossflow is not important for the conditions of

interest.
 2. HEAT TRANSFER IN DIRECT-TYPE HEAT EXCHANGERS

2.1 Dimensional and Nondimensional Variables

Heat exchanger design theory deals with various thermo-fluid relations

concerning the following two parameters :

• Heat transfer and fluid friction parameters.

• Principal design/operating parameters of the exchanger; such as flow

rates, flow distributions, thermo-physical properties of the fluids,

specified fluid temperature levers and exchanger effectiveness.

"
The parameters related to heat transfer characteristics are as follows:

U= overall heat transfer coefficient, W/K m 2

A = heat transfer area on which U is based, m 2

C= (m cp ) = capacity rate, W/K

m = mass flow rate, kg/s

cp= specific heat at constant pressure, J/kg K

gas (hot-fluid) terminal temperatures, K

T .
a.ln
air (cold-fluid) terminal temperatures, K

wherc subscripts a, g show air and gas rcspcctivcly.

5
 The importance of the parameters listed above is easy to compute, with

the exception of V. This term comes from an over~ll heat transfer rate equation,

which combines the convective and conductive mechanisms responsible for the

heat transfer from the hot to the cold fluid into a single equation similar to Ohm's

law for the steady-state flow of electrical current.

dq
dA
(2.1)

The term ~~ stands for heat flux per unit heat transfer area at a section where

the temperature difference is (Tg - T a ). The overall heat transfer coefficient of a

direct-type tubular air preheater, having three components (air-side convection,

gas-side convection, and wall conduction) may be expressed in the following forms;

1 (2.2)

1 (2.3)
;-ga 1-
hg + .!.E.-
kw In{ra/rg) + hIa

where rg and ra are the inside and outside radii of the tubes, respectively, and

k w is the thermal conductivity of the tube wall (\V 1m K).

Equations (2.2) and (2.3) define V in terms of the gas side heat transfer

area A g and air side heat transfer area A a , respectively. It is clear that equation

(2.4) relates Va and V g .

6
 (2.4)

The convective heat transfer coefficients h a and h g are complex functions

of the surface geometry, fluid properties, and flow conditions. They are

approximately calculated by experimental correlations.

The (-NTU theory groups heat transfer variables of a heat exchanger into

nondimensional parameters. These non dimensional parameters are named and

defined as follows

• heat exchanger effectiveness

Cg (Tg,in - Tg,out) C a (Ta,out

( (2.5)
Cmin ( Tg,in - Ta,in)

where Cmin is the smaller of the C g and Ca magnitudes. The quantity qmax IS

the maximum possible heat transfer rate for the exchanger. This heat transfer

rate could, III principle, be achieved in a counterflow heat exchanger of infinite

length. In such an exchanger, one of the fluids would expenence the maximum

possible temperature difference (Tg,in - Ta,in)'

• number of hEat transfe,- units

(.4 U)g (A U)a

NTU Cmin
(2.6)
Cmin

• capacity ,-ate ,-atia

(2.7)
where Cmin and C max are, respectively, the smaller and the larger of two

magnitudes Cg and Ca.

In general, it is possible to show that the heat exchanger effectiveness IS

explicitly dependent on:

1. the thermal size of the heat exchanger

2. the fluid flow heat capacity rate ratio

3. the flow arrangement in the heat exchanger

This can be shown with the following equation:

( = <p (NTU, CR , flow arrangement) (2.8)

Effectiveness expressions in this nondimensionalized form have been

developed for a variety of heat exchangers [1]. For instance, when a heat

exchanger is arranged as a cross-counterflow (single pass), and one fluid is mixed

and the other unmixed, the (-NTU relations are given as follows;

1. Cmax (mixed), Cmin (unmixed);

( (2.9)

2. Cmin (mixed), Cmax (unmixed);

( (2.10)
 Examination of Figure 2.1 demonstrates the asymptotic character of the f.

versus NTU relation for a given capacity rate ratio in a crossflow heat exchanger

where one fluid is mixed, the other unmixed [1].

On the other hand, there is no closed form effectiveness expression for a

crossflow exchanger, having both of the fluids unmixed. In this case the
------
effectiveness relation, in terms of non dimensional parameters CR and NTU, is

available in the series form proposed by Mason [2]. Mason's series solution was

further developed and suitably arranged for computer programing by Chung [3].

r--~---:--r-T-:T-;:;;-r-=::::::~7'"""~-i
,, + ~-
,
-!.-----l
I
,-----
~ "
-- t~
: \
---,-----------------~

----------i

1
E 1.33
0.75
,---2
~----~----------------~----_:_- - -0.5
( : . . 4 :
0.2 \- ---:-----~-----~-----~---.-~----.~-- ---------- 0.25 ·-----1
) .
---------,....- ------ --_ .... _- --- --- - -- "'-- - -- ---- - - ~ - ---~ - -- -. --- ---~

l I I ~

o'-- ~. ~ _
o 1 2 3 4 5
NTU

Figure 2.1 Effectiveness of a single-pass crossnow heat exchanger

with one of the fluids mixed and the other unmixed [20].

9
2.2 Assumptions Applied to f-NTU Theory

The f-NTU theory groups heat transfer variables of a heat exchanger into

nondimensional parameters. These nondimensional parameters are easily readable

and allow a compact graphical presentation. However, The f-NTU theory has

certain limitations because of the following idealizations:

steady state

constant and uniform velocity profiles at inlet sections

no longitudinal heat conduction in the flow direction;

in the solid wall

in the fluid

constant overall heat transfer coefficient

constant fluid specific heats

no heat loss to the sun'oundings

constant and uniform temperature profiles at inlet sections

no heat generation, no phase change occur in the fluid streams

flowing through the exchanger

uniform distribution of heat transfer area throughout exchanger core

2.3 Longitudinal Heat Conduction Effect. on Air Prehcat.er Performance

The (-NTU theory is based on the idealization that there is no longitudinal

conduction (in the flow direction). either in the solid wall or in the fluid. Fluids

generally have a low thermal conductivity (liquid metals excepted). but the wall

conductivity may be quite high for some cases.

10
 Many investigations about this effect have been performed for storage type

period-flow heat exchangers [1, 2, 4, 5], and for direct-transfer type heat

exchangers such as counterflow heat exchangers [6, 7, 8] and crossflow heat

exchangers having plate-fin surfaces [9]. Some of these investigations showed that

the deterioration of the exchanger effectiveness due to axial conduction is the

greatest when CR =l [1,4,7,9].

A complete study of this effect on direct-transfer type tubular crossflow

heat exchangers does not seem to be available in the literature. However, it is

stated that heat exchanger performance deterioration, because of longitudinal

conduction, is less significant for tubular heat exchangers than for plate-fin heat

exchangers [9].

A simple analysis was made III order to determine the amount of

deterioration occuring III the total heat transfer rate. In this study, only one tube

was considered and it was assumed that the air temperature was constant. Gas

flows through the tube while air flows around the outside of the tube,

perpendicular to the gas flow (see F ·IgU re C)C») Tlle d

_._. ·
eSlgn an d operating

parameters of the air preheater under consideration were the same a.s those of a

real tubular air preheater.

On the basis of these idealizations and design and operating conditions of

the air preheater, a set of differential equations describing the heat transfer

process for a typical element of the tube (see Figure 2.2) can be expressed as

follows:

• heat exchange b€tween gas and tube wall

(2.11)

11
 GAS

2r 9

...

z ...

L
... AIR

...

...
2r a

Figure 2.2 The dimensions of one tube of a crossflow tubular heat

exchanger. The longitudinal conduction effect was studied

using this particular tube (Chapter 2.3).

12
 ra+rg
where rm=--2-

• energy balance at the wall of the typical element

(2.12)

The boundary conditions are

T a = constant (2.13)

(2.14)

dJzm(O)=o (2.15)

The boundary conditions specify inlet gas temperature and mandate no heat

conduction at the inlet of the tube (Z=O).

The system of ordinary differential equations consisting of the equations

(2.11) to (2.14) can be written In nondimensional form by introducing the

following dimensionless parameters:

• dimensionless axial coordinate (in the gas flow direction)

(2.16)

• conduction pQ7"0771dcr

>. _"wAll' (2.11)

g- Leg

13
 • dimensionless temperatures; Om, Og, and Oa

0- =-T_-_T-:a;;.- (2.18)
-T . -T a
g,zn .

• number of heat transfer units on gas side

(2.19)

• number of heat transfer units on air side

(2.19)

Governing differential equations (2.11) and (2.12) become

dOg
d:: +o(Og-Om)=O (2.21 )

(2.22)

with the boundary conditions

(2.23) 0

d~;n(O)=O (2.24)

(2.25 )
The quantity ).9 is the conduction parameter which gives us valuable information

about longitudinal conduction effect on total heat transfer rate. When ).9=0, we

have no longitudinal conduction in the problem. The parameters Q' and f3 are the

number of heat transfer units on gas side and air side, respectively. The 8's are

the nondimensional temperatures (8 a =0).

The general solutions to eqns. (2.11) and (2.12) are

(2.26)

(2.27)

where the ,'s represent the roots of the characteristic equation;

(2.28)

The roots of the above equation were numerically computed by Newton's method

and they are in the following ranges;

'I <0. and (2.29)

Since the solution is exponential. the root '2 causes the solution to go to

infinity. Therefore, c2 and b 2 must have zero values because the temperature

profiles don't go infinity in an actual heat exchangers.

The rest of the constants c 1 ' c 3 . b 1 • and b 3 may be found froIll the

boundary conditions. eqns. (2.23) and (2.24);

15
 (2.30)

(2.31)

(2.32)

(2.33)

vVhen there is no longitudinal conduction p. g =0), the solution of the

governing equations is greatly simplified. It is

-a{l-_l } z
0:+(3
e (2.34)

-O:{l-_l } z
0:+(3
Bm = [a~(3] e (2.35)

The results of this study are presented in Figure 2.3. \\Then comparing

nondimensional gas temperatures, one with longitudinal conduction and the other

without, we find that they are identical up to 4 decimal places. The same goes for

tube wall nondimensional temperatures. ~Ioreover. the difference between the

total heat transfer rate of each case is less than 0.0005%. It is quite obvious that

longitudinal heat conduction has a negligible effect on heat transfer under normal

design and operation conditions.

16
 1.~I!I---""---,--.----,----------------'
--0- e
g: without axial conduction
• eg: with axial conduction
-0- em: without axial conduction
~ 0.8
~ • em: with axial conduction

0.6

r:
c o.

c 0.2
Z

0.0 L . . - _........_ - J ._ _........_ - l . ._ ___''''---_..l...._--'-_ _L-_--'-_----l

0.0 0.2 0.4 0.6 0.8 1 .0

---...:;>~ Z (Z/L)

Figure 2.3 l'\ondimensional temperature profiles along the one tube of

the crossnow exchanger: the two with longitudinal heat

conduction P g =0.00031). Eqns 2.26 and 2.27. the other

two without Pg=O.O). Eqns 2.3·1 and 2.:3.').

17
 3. CONVECTION CORRELATIONS

3.1 Introduction

The emphasis in this section is placed on forced convection correlations for

flows inside circular tubes and over tube bundles and their applicable ranges. The

correlations presented here are limited to steady, incompressible flow of constant-

property Newtonian fluids. Any effects of natural convection, phase change, mass

transfer, chemical reactions, thermal energy sources, viscous dissipation (i.e.,

internal friction), flow work (i.e., work done by pressure forces), and fluid axial

conduction are omitted. Moreover, the tube walls are considered to be smooth,

rigid, and stationary.

3.2 Flow Inside Circular Tubes

Laminar, transition, and turbulent flows and heat transfer characteristics

of circular tubes have been studied in great detail, as this geometry finds

widespread use 10 practical applications. Different investigators performed

extensive experimental and theoretical studies with various fluids. As a result,

they formulated relations for the Nusselt number vs. the Reynolds and Prandtl

numbers for a wide range of these nondimensional groups.

There are four types of flow in circular tubes, namely, fully developed.

hydrodynamically developing, thermally developing and simultaneously

developing. Prandtl number is the key parameter in this classification since it is

the ratio of the momentum diffusivity v to the thermal diffusivity o.

18
 In the simultaneously developing flow case, the viscous and thermal effects

diffuse simultaneously from the tube wall, commepcing at the tube entrance. This

happens when Prandtl number is near unity, which includes gases such as flue gas

and aIr.

Constant heat flux is often considered as a thermal boundary condition

imposed on a tube wall for gas-to-gas heat exchanger applications. This is

because the average heat flux of a row of a tube bank increases until

approximately the fifth row, after which there is little change in the turbulence

and hence in the average heat flux.

3.2.1 Laminar Flow

Numerous results are available for forced convection laminar flow of a

circular tube in the literature. These results have been compiled in a monograph

by Shah and London [10] and in an updated review by Shah and Bhatti [11].

For simultaneously developing laminar flow in a circular tube with

constant heat flux boundary condition, Heaton, Reynolds, and Kays [12] obtained

a solution by an integral method. Their result showed excellent agreement with

experimental measurements made for Prandtl number=O.7. Their tabulated

Nusselt numbers are used for the present analysis (Table 3.1). The Nusselt

number for fully developed laminar flow in a circular tube is constant (Nu=4.3G).

Experimental studies by Kays [14] showed that a.s long as the values of

(x/D) range from 48 to 80 there is no measurable effect on the Nusselt numbers

when comparing different entry types (Pr= O. 7) for the laminar flow of a circular

tube. However, the behaviour of very short tubes might be different.

19
 Table 3.1 Results of the heat transfer analysis for simultaneously

developing laminar flow inside a circular tube (Pr=0.7) [13]

=======================================

x*{Re.Pr/(x/D)}

=======================================

0.00010 51.90

0.0010 17.84
~
0.0025 12.08

0.0050 9.12

0.010 7.14

0.025 5.49

0.050 4.72

0.10 4.41

0.25 4.36

=======================================

3.2.2 Turbulent Flow

Turbulent circular tube flows have immense technological importance, as

they occur frequently under normal operating conditions for a variety of heating

and cooling devices.

The observations show that a laminar flow pattern transforms to a chaotic

turbulent flow pattern when the Reynolds number exceeds a certain critical value

called the critical Reynolds number Re crit ' In the case of a fully de\'c\oped flow in a

circular tube. the 100ver limit of Re crit is accepted to be 2300. whereas the highest

20
value of the upper limit attained by Pfenninger [11] is 1.001 x 10 5 . Although the

upper limit of Recrit is undefined, for most practical purposes the flow in the range

2300 ~ ReD ~ 10 4 is regarded as a transition flow region. Compact heat exchangers

at part loads may operate in the transition region although the design value of

ReD falls above 104

Extensive efforts have been made to obtain empirical correlations that

either represent a best fit curve to the experimental data or have the constants in

the theoretical equations adjusted to best fit the experimental data. An example

of the latter is the correlation given by Petukhov and Popov [14]. Their

theoretical calculations for the case of fully developed turbulent flow with constant

heat flux boundary condition yielded the following correlation, which is based on

the three-layer turbulent boundary-layer model with constants adjusted to match

the experimental data:

(f/8) Re D Pr
N ufd = ---------'----'--=--.-1----=-1---,2.---- (3.1)

(1 +3.4f) +( 11. 7 + 1.8Pr -3)(f/S)2( Pr3 -1)

where

(3.2)

is the friction factor (r 1

Spum
TW 2)'
Equation (3.2) is applicable for fully dcvclopcd turbulent flo\\' III thc rangc

21
104~Reo~5 x 105 and 0.5<Pr<2000 with 1 % error [14]. A simple correlation
has also been proposed by Petukhov, Krillov, and Popov [15] as

(f/8) ReOPr
NUfd = 1 2 (3.3)
1.07 + 12. 7( f/8)2( Pr3 -1)

where f may be obtained from the Moody diagram or, for smooth tubes, from

eqn. (3.2) The above correlation (Eqn. 3.3) predicts the results in the range

104<ReO<5 x 10 5 and 0.5<Pr<200 with 5 to 6 % error [15].

There are numerous heat transfer correlations which have been established

for fully developed flow in a circular tube. A compilation of such correlations has

been summarized by Kays and Perkins [17], by Shah and Bhatti [11]. One

correlation, which is widely used and is attributed to Gnielinski [16], is of the form

(f/8)( ReO -1000)Pr

NUfd = 1 2 (3.4 )
1+12.7(f/S)2(Pr3 -1)

where, for smooth tubes, the friction factor is given by

f={ O. 79In[ Re O] -1.64}-2 (3.5 )

The Gnielinski correlation is a modified version oJ the second Petukhov et al.

correlation (Eqn. 3.3), extending it to the 2300<ReO<5 x 10 4 range. For

0.5:SPr:S2000 and 2300:SReO:S5 x 10 6 , it is in overall best accord with the

experimental data; it agrees with the second Petukhov et al. correlation within
-2% and +7.8%. Hence it is selected as the common basis of comparison for all

the correlations in [11]. The Gnielinski correlation along with the friction factor

given by equation (3.5) was used to calculate Nusselt numbers in the present

analysis of an air preheater.

All correlations presented so far apply for both uniform surface heat flux

and temperature. Properties should be evaluated at an average bulk temperature.

Although entry lengths for turbulent flow are typically short,

10::;(xfd/D)~60, it IS often preferred to consider the entry effect for a

simultaneously developing flow. Mills [18] carried out extensive experimental

investigations to study this effect in a smooth circular tube using air (Pr=O.7) as

the working fluid, employing the constant heat flux boundary condition. The

mean Nusselt number NUm is expressed for the five entrance configurations shown

in Table 3.2 by formulas of the type

(3.6)

where Nu fdstands for the fully developed Nusselt number. The constants C and n

depend on the nature of the inlet (for example, sharp-edged or nozzle) and entry

region (thermal or simultaneously developing). as well as on the Prandtl and

Reynolds numbers.

From eqn. (3.6). the mean Nusselt number between any two arbitrary

points along the tube might be expressed as in the following equation

(3.7)

n
Table 3.2 Ratio of 1Vlean to Fully Developed Turbulent Flow Nusselt

Number in the Entrance Region of a Smooth Circular Tube

with Various Entrance Configurations for Pr=O.7 [18].

En trance Configuration Formula for Num/Nu""

Adiabatic surface q~

/I////'i>r////lII//// ~llW
NU m 0.9756
--=1+----
Nu oc (x/D h )0760
\\\\\\\\\\\\\\\\\\\t~

Long calming section

Square entrance

Nu., 09759
= 1 +
Nux (x/D h )0700

180 Round bend

0

10517
= 1 4-
(.tID" )06)0

90 Round bend
0

90- Elbow
3.2.3 The Influence of Surface Roughness

All the preceding discussion has been based on the assumption of an

aerodynamically smooth surface. Circular tube wall roughness has little effect on

laminar flow. However, it exerts a strong influence on turbulent flow.

Surface roughness can take many forms. The name of Nikuradse [19] is

indelibly associated with the rational analysis of rough surfaces. He performed

systematic experiments with sand grains glued onto the interior of circular tubes.

The symbol k, a length dimension, is used to describe the roughness element size,

k being actually the size of the sieve used by Nikuradse to sift the sand.

The effect of roughness on a turbulent boundary layer is felt mostly right

at the wall, and thus a nondimensional expression of roughness size is logically

based on the shear velocity UT defined as

(3.8)

This leads to a roughness Reynolds number Re k as a nondimensional measure of

surface roughness:

u-k
Rek=+ (3.9)

\Vith this roughness Reynolds number ReI. as the parameter, Nikuradse identified

the following three flow regimes depending on the variation of f with ReI. and
 1. Hydraulically smooth regime, 0~Rek~5: f f(Re D )

2. Transition regime, 5~Rek~70: f=f(k/r g , ReD)

3. Completely rough regime, Rek>70: f=f(k/rg)

The roughness used by Nikuradse in his experiments doesn't represent the

type of roughness encountered III commercial circular tube surfaces. To

circumvent this difficulty, Schlichting [20] introduced the concept of equivalent

sand-grain roughness for roughness elements such as spheres, spherical segments,

cones, and short triangles. Moody [21] determined the equivalent sand grain

roughness for eight types of commercially available circular tube surfaces. His

results, some of which are tabulated in Table 3.3, are very useful in practical

applications.

Table 3.3 Equivalent sand roughness for commercial circular tube

su rfaces [21].

Type of the circular tube

riveted steel 900-9000

reinforced concrete 300-3000

cast iron 240

commercial steel 46

26
 '~l

0.1 II
I Critical I II
I I IJ
0.09
0.08
"
,II . I
Laminar
zone
Transitidn~
l!ow - zone Fully rough zone
0.05
0.07 " 0.04
'\
0.06 - 0.03
0.05 ~~~' 0.02
-. ~
Q
0.015
~l~'
n. 0.04CIE
\\ ~\
~
too..
r----
to--.. ----
f-",
~

0.01
-!,!\C).

'U
1Cl.
" '7J\C1'
'" ~
\ " 0.008 tfi
VI
l::l -
,.......... 1'-0 -' I'· 0.006 QJ
C
~, r---..
"
' ...... 0.03 1\
0.004
till
:J
o
~
...: I-

.....
0
f-
QJ
>
U 0,025 :;::;
I~ ~ 0.002 ro
-1 C
r--. r--.
Q3
0 \ ..... 0::
'';::;
u 0.02 ::::---...:: r- I-... - 0.001
........ r--.
';:
lJ...
~, 0.0008
~ ....... 0.0006
~
~ I- 0.0004
0.015 ReD c Smooth pipes-
r-- I, 0.0002
I
I II /,; Cum) r----..... r--. r....
0.0001
Drawn tubing 1.5 f"': ~ 0.000,05
0.01
Commercial steel 46 "" f::::::t'-
Cast iron 260
0.009 Concrete 300-3000 t - to-
0.008 t-.... 0.000,01
1cP 2 3 4 56 810 4 2 3 4 56 8 105 2 3 4 56 810 6 2' 3 456 8107 ....... ~~\5~108

.k
D
= 0.000,001 \.k
D
= 0.000,005
Reynolds number, ReD = - vm D-
II

Figure 3.1 Friction Factor for Fully Developed Flow in a Circular Tube [21].
 We now turn to the effect of roughness on the heat transfer rate. Two

distinct influences of the roughness elements are recognized. First, they increase

the tube surface area, ~nd second, they increase the heat transfer coefficient. This

latter effect is brought about by the change in the turbulence patterns close to the

wall. The following simple correlation, suggested by Norris [22], expresses the

effect of roughness for turbulent flow in a circular tube:

for £~4 (3.10)

where n=0.68PrO. 215 for 1 <Pr<6. For £ >4, Norris observed that the Nusselt

number no longer increases.

Moody's friction factor plot given III Figure 3.1 may be used to determine

the effect of surface roughness on the heat transfer rate. For example, consider

the following conditions which are some of the operating and design conditions of

the air preheater under consideration:

ReD= 30,000 (turbulent flow in the circular tubes)

Pr=0.7 (flue gas)

k/ D=0.001 (commercial steel tube with an inside diameter of 0.0146 m)

Using Moody's diagram and Norris's correlation, it was found that

Is=0.024

1=0.027

28
 Nu =1.077
N Us

This shows that heat transfer rate Increases by 7.7% due to surface roughness for

conditions of interest.

3.3 Flow Over Tu be Ban ks

3.3.1 Banks of Tubes

In practice, air heaters made from banks of tubes are widely in use. The

tubes in a bank are often arranged in staggered or in-line configurations. Fig. 3.2

illustrates these two arrangements and definitive characteristic dimensions of a

tube bank

a=ST/D=the relative transverse pitch (3.11)

b= SLID = the relative longitudinal pitch (3.12)

c= SD/D = the realtive diagonal pitch (3.13)

Flow conditions within the bank are dominated by boundary layer

separation effects and by wake interactions, which in turn influence convection

heat transfer. The majority of experimental investigations showed that the heat

transfer in a tube in a bank is greater than that of a single tube and depends on

longitudinal and transverse pitches. From the heat transfer standpoint, the

staggered arrangement is more effective [ 23 ].

Banks of tubes act as \"ortex generators, and. depending on the location of

29
the tubes, a corresponding turbulence level is established in each bank. Therefore,

heat transfer for tubes in inner rows is considerably higher than heat transfer for

tubes in the front rows [ 23 ].

a) b)

Figure 3.2 Tube arrangements III a tube bank; (a) in-line, (b) staggered.

3.3.2 Mean Heat Transfer from Banks of Tubes

Heat transfer of a tube in a bank depends mainly on the velocity of the

thermal carrier, tube arrangements, properties of the incoming

, fluid, thermal

loading, turbulence level, and direction of heat flux.

Experimental results of heat transfer for banks of tu bcs in gas flow wcrc

generalized according to similarity theory and wcrc cxpresscd by t hc cq U<I tion of

similarity (NuD=CRemDmaxPrn). Zukauskas [23] has proposed a correlation of

30
the form

Nu =CRe m PrO. 36 ( Pr )0.25 (3.14)

o Dmax Prs

This correlation is applicable to both staggered and in-line tube arrengements with

=f 15% error, having the following restrictions

N L =row number of tube banks>20

0.7<Pr<500

All properties except Prs are evaluated at Ta,in (inlet bulk temperature of the

fluid). The constants C and m are listed in Table 3.3. Reynolds and Nusselt

num bers are defined as follows

Re _PYmaxD (3.15)
Dmax - /1

(3.16)

where 0 is the outside diameter of the tubes (2r a ).

Since the Prandtl number is a.lmost constant for gases, pPr :::::1. The Reynolds
rs
number ReO max is based on the maximum a ....erage fluid velocity
• occuring within

the tube bank. For the in-line arrangement. V max occurs at the trans ....erse plane

31
Al of Figure 3.2a, and from mass conservation for an incompressible fluid:

Vmax=~l
a- V
(3.17)

Table 3.4 Constants of Equation 3.14 for tube banks in crossflow [23].

=========================================

CONFIGURATION ReOmax C m

=========================================

In-line 10-10 2 0.80 0.40

Staggered 10-10 2 0.90 0.40

In-line 10 2 -10 3 0.52 0.50

Staggered 10 2 _10 3 0.52 0.50

In-line 10 3 -2x10 5 0.27 0.63

CST / SL <0.7)*

Staggered 10 3 -2x10 5 0.35C~T)0.25 0.60

L
CST/SL <2)

Staggered 10 3 -2x10 5 0.40 0.60

CS T / SL>2)

In-line 2x10 5 -2x10 6 0.021 0.84

Staggered 2xl0 5 -2xl0 6 0.022 0.84

=========================================

* For ST/SL>0.7, heat transfer is inefficient and in-line tubes should not be used.

For a staggered configuration, the maximum average velocity may occur

32
at either the transverse plane Al or the diagonal plane A 2 of Figure 3.2b. It will

occur at plane Al if the rows are spaced such that

2(c-l)«a-l) (3.18)

The factor of 2 results from the bifurcation experienced by the fluid moving from

the Al to the A 2 planes. Hence V max occurs at A 2 if

(3.19)

in which case it is given by

(3.20)

If V max occurs at Al for the staggered configuration, it may again be computed

from equation 3.14 [24].

Flow around tubes In the first row of a tube bank corresponds to that for

a single (isolated) cylinder In crossflow. However, for subsequent rows, the flow

depends strongly on t.he tube bank arrangement (Figure 3.3). In-line tubes

beyond the first ro\V are in the turbulent wakcs of upst.ream tubcs, and for

modera t.e values of SL' convcct.ion coefficien t.s associated wit h downst ream rows

are cnhanced by t.urbulation of the flow. Typically. the con\'cction coefficient of a

row increases with increasing row number until approximat.ely thc fifth row, after

which there is little change in t.he turbulence and hence in the convection

33
coefficient. However, for small values of SL' upstream rows, in effect, shield

downstream rows from much of the flow, and heat transfer is adversely affected.

That is, the preferred flow path is in lanes between the tubes and much of the

tube surface is not exposed to the main flow. For this reason, operation of in-line

tube banks with ST/SL>0.7 (Table 3.4) is undesirable. For the staggered array,

however, the path of the main flow is more tortuous and a greater portion of the

surface area of the downstream tubes remains in this path. In general, heat

transfer enhancement is favored by the more tortuous flow of a staggered

arrangement, particularly for small Reynolds number (ReD<lOO) [24].

From ReDmax equal to 1000 and above, the heat transfer of the first rows

of the tubes decreases further, compared with heat transfer of the inner rows.

This is an indication of how turbulence, generated by the first rows of tubes in a

bank, affects heat transfer of the mner rows. At ReDmax equal to 1000, the

difference in heat transfer of the first and inner rows amounts to 24%. At

ReD max equal to 10 5 , this difference rises to 70% [23].

3.3.3 The Influence of Roughness on the Mean Heat Transfer of a Bank

Heat transfer of smooth tube banks in crossnow has been extensively

investigated [23]. However, the effect of surface roughness on the heat transfer

rate in those investigations has not been sufficiently studied.

Zukauskas [23] proposed correlations for the calculation of the mean heat

transfer coefficient in rough tube banks on the basis of other investigators'

experimental results for local heat transfer coefficients. \Vhen Re D711ax ranges

from 10 3 to 10 5 , the correlation is

34
 (3.25)

and in the Re Dmax range from 10

5 to 2xl0 6 , the correlation is

(3.26)

The mean heat transfer can be computed from these equations with an accuracy

of ± 15% for a in the range from 1.25 to 2.0, b from 0.935 to 2.0, and k/ D from

6.67xlO- 3 to 40xlO- 3 .

In the flow of air over a tube bank, the effect of roughness is manifest at

higher ReDmax (ReDmax=5xl04 to 7xl0 4 ) for the k/ D range from 0.3xlO- 3 to

8xlO- 3 [23]. Since k/ D is near lxlO- 3 and Re Dmax changes from 2xl0 4 to 5xl0 4

in the present analysis of the air preheater, the correlations for rough tube banks

proposed by Zukuaskas were not used for calculating the mean heat transfer rate.

Instead all analysis based on the correlations for smooth tube banks proposed by

Zukuaskas [23].

35
 --t>
~
0
~- C' fd ;~u '\ ~- u ----------
~---l----
--t>
-{> 0 \. U 0
~- / l
-
__ t ,
-
0 r) C
\.
~-:> '-- flow
Preferred
lanes

--t> --
~L
---------------~
-
'-

-
0
----c...:>
----
--......
- - - - -
---{>
-l----
'

---{>
- Oe--:;--O ceO:=--=--
~

~
~5
~

~~
~ ?
~

C
----------

(a)

(b)

Figure 3.3 Flo\\" conditions in a bank; (a) in-line. (b) staggered.

36
 4. MATHEMATICAL MODEL

4.1 Single-Pass Crossflow Exchanger

The mathematical model is based on one of the two basic heat exchanger

theories which has been used extensively; namely the (-NTU method [1]. In the

heat exchanger shown in Figure 4.1, flue gas flows through the tubes while air

flows over the tube bundle.

The overall heat exchanger is divided into smaller crossflow heat exchanger

assemblies in the gas flow direction, i, and in the air flow direction, j (see Figure

4.1). A small element may consist of a number of tubes, depending on its sIze.

For modeling purposes, each small element of the heat exchanger was

approximated by a crossflow heat exchanger, with one of the fluids (air) mixed

and the other (gas) unmixed. It has also been assumed that the mass flow rates

at each row i, and column j, don't change throughout the flow length. That is,

both fluids are flowing in closed passages, representing .. he rows and columns In

Figure 4.1. Moreover, these smaller heat exchanger assem blies are considered to

have uniform metal temperatures since their dimensions are small compared to

the overall heat exchanger.

Using the (-NTU Method [1], outlet gas temperature, outlet aIr

temperature and the tube metal temperature of each element ha\'e been

calculated. Considering the ij rh element in Fig. 4.1 and using the equations

presented in Chapter 2.1,

37
 GAS x

Tg 1.J !,
Tm 1.1
i
DY(i) To l.j Tm lJ To 1.J +1 Tm 1.N
~.

: . -D X(j)--' Tg 2.J

Tg I.J

0
Tg 1.1
AIR
Tm 1.1 Tm I.N
101.1
To 1.1 To 1.2 101
-1<1

Tg 1+ 1.1
Tg 1+ l.J

Tg M.l Tg M.N

To M.l Tm M.l ToM2 Tm M.j To M.N Tm M.N To M.N+1

Tg M+1.1 Tg M+ IN

Figure 4.1 Subdivisions of a single-pass crossflow heat exchanger

on xy plane.

38
the capacity rate ratio of the ij th element is

C ...
min IJ
CR r=c
fj
.:
max IJ
(4.1)

The number of heat transfer units of the ij th element is

(AU) ..
NTU .. gIJ
IJ Cmin
...IJ ' (4.2)

If C .. is mixed and C . .. unmixed;

max IJ min IJ

the effectiveness of the ij th element becomes

( .. =(1/C R .. ){l-exp[l-exp(-NTU .. )J}, (4.3)

IJ IJ IJ

The outlet air and gas temperatures of the ij th element are

T alJ.. = T alJ"+1-( IJ"C R IJ.. (TglJ.. - T alJ"+1) (4.4 )

Tg'+1'=
l) TglJ.. -( I).. ( T glJ.. - T alJ"+1)' (4.5 )

If C . .. is mixed and C .. unmixed;

min IJ max IJ

the effectiveness of the ij th element becomes

( .. =l-exP(-[l/C R ..;{l-ex P(-C R .. NTU .. )}), (4.6)

IJ IJ IJ IJ

and the outlet air and gas temperatures of the ij th clement are

39
 T al).. = T al)"+1-( I):.( Tgl).. - T al)"+1) (4.7)

T gl·+1'=
J T gl).. -( I)"C R I).. ( T gl).. - T al)"+1)· (4.8)

Total heat transfer rate of the ij th element may be computed from

Q=(mc p ) al).. ( T al).. - T al)"+l)=(mc p ) gl).. ( T gl).. - T gl'+1')

J (4.9)

or in terms of the tube metal and average gas temperatures,

Q_ A g I)··({fTg I).. + T g "+11/2} -

I)
T m I).. )
(4.10)
- {1 rg
~+k In(rm/rg}
}
g I) W

Equating eqns 4.9 and 4.10, we find the tube metal temperature of the I) th

element, T .. as
ml)

(T .. - T .. 1) r
T m I).. =(mc p }a' I). a I)
A a
..
1)+ {1 g
-h-+-k In(rm/rg} }
.. W
- T gavg I).. (4.11)
gl) gl)

where T .. is the average gas temperature of the ij th element.

gavg I)

Knowing inlet gas and aIr temperatures, we are able to directly calculate

the tube metal temperature and outlet gas and air temperatures of the ij th

clement. Now we should find a computing sequence so that the same calculation

can be done for all clements. This may be accomplished with a dou ble do-loop in

40
the computer program, commencing at the right hand-side upper corner element.

That element is the (1,N) th element 1ll Fig. 4.1. In this case, the temperature

and velocity profiles at inlet sections of the overall heat exchanger must be known.

It is obvious from Figure 4.1 that calculations must proceed column by column or

row by row. Having evaluated the temperature fields of the overall heat

exchanger, we may calculate the overall effectiveness by equation 2.5, using

average inlet and outlet temperatures of the working fluids.

Or

4.2 Verification of the Model

The mathematical model presented in Chapter 4.1 enables us to predict

the performance of a single-pass crossflow heat exchanger having both of the fluids

unmixed.

The accuracy of the results of the mathematical model depends on the

number of subdivisions (elements) used. The use of more subdivisions can

produce higher accuracy. Practically, the elements used must be determined

according to the accuracy desired.

Using a number of different sIze matrices, the effectiveness of a single-pass

crossflow exchanger having both of the fluids unmixed, for the case of C =O.8103
R
and NTU = 1.8632, was calculated. The results are presented in Figure 4.2. In

this calculation, the following parameters were set to a constant value in order to

make a comparison with the analytical solution of the same case (C R =O.8103 and

NTU = 1.8632) [3];

1. inlet gas temperature and velocity

2. inlet air temperature and velocity

·11
 3. physical properties of working fluids

4. overall heat transfer coefficient

It can be seen from Figures 4.2 to 4.5 that as the number of subdivisions

increases, the accuracy of the mathematical model results increases as well. When

a 10xlO matrix IS used, these results are accurate to the 3rd digit for most cases,

as compared with the analytical solution [3]. Three digit accuracy is believed to

be sufficient for most engineering applications. Thus, a 10xlO matrix arrangement

was used throughout this investigation.

In Table 4.1, the comparison between the results of the presented

mathematical model and the analytical solution is extended to various cases of

capacity rate ratio (C R ) and number of heat transfer units (NTU).

As seen from Table 4.1, the effectiveness of the present computer model

deviates by 0.05% from the analytical solution by Chung [3], that is, we have

agreement to within three digits for given values of C R and NTU. Using another

approach, Chiou [9] also came up with three digit agreement in the effectiveness

by solving the governing heat transfer equations for a crossflow exchanger [2] with

the aid of a finite difference mathematical method. But computationally, this is

relatively expensive compared to approach used in the present analysis.

4.3 Two-Pa.ss Counter-Crossflo\\' Air Preheater

The objective of this work was to find the metal temperature field of a

tubular recuperative air preheater arranged as a two-pass counter-crossflmv (see

Figure 4.4). The inlet gas and air temperatures of the air preheater and the inlet

velocity variations of air and gas are given as boundary conditions.

 Element Size Effect on the Heat Exchanger Effectiveness

66 . . • . . •
~

--
0~
65
~
~
- Both fluids unmixed
...
CD
,.
.
C)
'llc
.
cu • • • • •
.c
Co>
64 - • -
>< •
-
CD

cu • CR = 0.8103
CD
.c 63 - NTU = 1.8632

-0

U)
-
U) 62 I-
CD
c
CD

--
>
-
One fluid mixed. the other unmixed

-
Co>
CD 61
w

I I
60
1 10 100 1000

Number of Subdivisions (MxN)

Figure 4.2 The effect of grid size on the heat exchanger effectiveness.

43
 .1
'-'
I 8
8 II
II ~
~ x
x"" It
_0<:;
-2:'-'
'-' .01
.
~.
I1
II j
Il __
..-.... ..................... ..
.001
.- l
I: l:
,~
I. l
l
!
I u j

I
.0001 _...- ................... ...........

•
; ~~

I
lI

II .-
I~ Il
I
j
o
o 200 400 600 800 1 000

subdivision nurnher (l\IxN)

Figu re 4.3 The effect of grid size on the accuracy of the mathematical

model results for heat exchanger effecti\·cness.

·H
 Element ISJze Effect on the Total Heat Transfer Rate
X 101\3)

107.50

107.25 .oJ"- -
-
,,/
-
3: ~
-
.:t:. 107.00

-CI)

aJ
a: 106.75
CR = 0.8103
-
I-
CI)

U)
l:
aJ
l-
106.50
NTU = 1.8632
I--

-aJ
CI)
106.25
J:
106.00
-aJ
0
I--
I
105.75 u

105.50
i
o 100 200 300 400 500 600 700 800 900
Number of Subdivisions (MxN)

Figure 4.4 The effect grid size on the total heat transfer rate.

45
 .1

I,
I .
I 8 I,
8 II
II l
l
II ~ 1
<: ~ .01
Ii
x oc::
""':-~
--:: '-'
CY[ I
•
~ j
l
1
I
.001
I
II

I
I .
• Ii
!
;
Iii
j
.0001
• I, ~

a
!
!
i
II
!
• !
I
1
~
a 200 400 600 800 1000

suhdivision number (r-.1xN)

Figure 4.5 The effect of grid size on the accuracy of the mathematical

model results for heat transfer rate of the heat exchanger.

Table 4.1 Comparison between the effectiveness of the mathematical model

presented in Chapter 4.1 and the analytical solution [3] for a

single-pass crossflow exchanger having both fluids unmixed.

-
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
-

NTU f analytical [3]

--
- --
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
----
-

2.042 0.6884 0.6885

2.849 0.7548 0.7551

3.377 0.7850 0.7854

3.490 0.7905 0.7910

3.518 0.7919 0.7923

C R =0.80 1.828 0.6391 0.6392

2.512 0.7016 0.7019

2.943 0.7299 0.7303

2.993 0.7328 0.7332

3.045 0.7357 0.7361

-C =0.90 1.677 0.6006 0.6008

R
2.278 0.6594 0.6597

2.647 0.6857 0.6861

2.687 0.6883 0.6886

2.725 0.6906 0.6910

===========================================

·17
 To accomplish this goal, we simultaneously employed the mathematical

method presented in Chapter 4.1 to both passes of the air preheater using an

iterative procedure proposed by Korst [25]. It has also been assumed that there is

adiabatic mixing inside the duct which connects the passes of the air preheater.

Owing to perfect mixing and n~ heat loss to the surroundings from the duct, the

gas will have a constant temperature throughout the duct (at <2>, <3>, <4>

in Figure 4.6). That is, the gas temperature at the inJet of Pass I (at <4» will

be the same as the average gas temperature at the outlet of Pass II (at <2».

Consider the sketch or-the air preheater in Figure 4.6 and the numbers

inside the angle brackets which reference locations III the sketch. The inlet gas

temperature of Pass I «4» is assumed. With this assumed gas temperature

entering Pass I «4», and the known inlet temperature profile of air «6», the

mathematical model for the single-pass crossnow case may be applied to Pass I in

order to obtain the temperature profile of air leaving Pass I «7» and entering

Pass II «7». \Vith the calculated inlet air temperature profile «7» and

known inlet gas temperature distribution «1», the same procedure is applied to

Pass II. Since the average gas temperature at the outlet of Pass II «2» is

carried out to Pass I as the inlet gas temperature «4», one iterative calculation

has now been completed. The same procedure is followed with the last calculated

value of the inlet gas temperature of Pass I until convergence to a solution is

reached.

A similar iterative procedure was used by Stevens [26] III order to find the

effectivenesses of two and three-pass crossnow exchangers.

48
 GAS IN GAS OUT

• <1>. x t <5> t
411 ....
AIR '-<8> PASS II PASS I
...- AIR
<7> <6> IN
OJT
411 ....
Y
<2> <4>

L. <3> --l

Figure 4.6 The sketch of the tubular recuperative air preheater

aITanged as a two-pass counter-crossnow.

Since the solution procedure of the computer model IS iterative. a

convergence criterion is needed. It is assumed that convergence -is achieved when

the average difference between the total heat transfer rates of the air preheater

calculated in two subsequent iterations is within an acceptable error. The average

difference in the total heat transfer rate is defined as

49
 X- Qnew- Q old
(4.12)
- Qnew

where the subscripts 'new' and 'old' stand for the current and the previous

iterations, respectively. When the convergence parameter becomes

(4.13)

the iteration ends. The convergence parameter X decays exponentially after 20 to

40 iterations, depending on the initial guess of the gas temperature at <4> in

Figure 4.6 and the capacity rate ratio and the number of heat transfer units of the

air preheater. Examination of the criterion, X, showed that convergence to the

solution of the problem is reached by the following form of an exponential

function,

X=exp{ -a ( number of iterations )+b} (4.14)

where constants a and b change for different operating conditions of the air

preheater.

Figure 4.7 shows the representation of Pass J of the air preheater. As seen

from this figure, Pass J is the upside-down mirror image of Figure 4.1 which

represents Pass II of the air preheater. Thus. the equations represented in

Chapter 4.1 are still applicable to Pass J.

50
 Tg M+ 1.1 Tg M+l.N

To M.l Tm M.l To M.2 Tm M.] To M.N Tm M.N To M.N+l

Tg M.l Tg M.N

Tg 1+ 1.1 Tg 1+ I.J
AIR

Tm L] ., Tm LN
To 1.1 Tm Ll To 1.2 To I.J To I.J +1

Tg 1.1 Tg I.J

~DXU)----.
Tg 2.]

Tm 1.1 Dl) To l.J Tm 1.] To l.J +1

~
Tm 1.N

Tg 1.]

GAS

Figure 4.7 Subdivisions of Pass I of the Air Preheater on xy Plane

51
5. THE EFFECT OF AIR LEAKAGE ON AIR PREHEATER PERFORMANCE

5.1 Air Leakage Model

The tubes III the air preheater are expanded into tube sheets at both ends.

To provide for expansion, one tube sheet should be free to move with respect to

the casing. Accomplishing this IS not possible without air leakage. Due to the

difference in pressure between the air and gas streams, some air leaks into the gas

stream through the tu be sheets at the cold and hot end of the air preheater. The

rates of air leakage increase as the air preheater stays in service due to a

loosening of the joints and because of acid corrosion.

The air leakage has a negative effect on air preheater performance. An air

leakage model has been developed for the two-pass counter-crossflow tubular air

preheater in order to predict the effect of the air leakage on the performance of

this particular air preheater (see Figure 5.1 and 5.2). It was assumed that air

leakage occurs at the inlet and outlet sections of both passes (see Figure 5.1).

Also, the leakage calculat.ions a.re based on average t.emperat.ure and mass flow

rates of the working fluids; therefore, the velocity and t.emperat.ure profiles

remained unchanged at. t.he locations where adiabatic mixing between the main

gas flows and leakage flows occur (see Figure 5.1).

 The leakage coefficients are defined as follows:

• cold end leakage coefficient on either pass

mL GE
°L GE= e' ( 5.1)
, m gi2

where

mL , GE=mL ,cet.+mL ,ceo (5.3)

• hot end leakage coefficient on either pass

mL,HE
oL , HE=-"e--
m gi2
(5.2)

where

m L,H E = m L,hei + m L,heo (5.4 )

The total aIr leakage of either pass is the sum of the cold and hot end leakage,

therefore, the total air leakage coefficien ts of Pass I and Pass II, respectively, are

_ m _
oL 1 - - Ll
e - - oL
711·')'
GEl +f, L • HE1 (5.5 )
g7_

L711 ')
oL_,)=-,---=0
me. L,GE:!. +f, L,H E:!. ( 5.6)
g12

There are two different leakage flows 10 opposite directions. at both the cold and

53
hot ends. One of the leakage flows is through the gas-out section and the other is

through the gas-in section (see Figure 5.1). The ratio between these two leakages

is denoted by R CE and R HE at the cold and hot end, respectively.

R _ mL,ceo
(5.7)
CE- mL , cei

R _ mL,heo
HE-m L , h el. (5.8)

Another ratio parameter is defined such that,

b
R _ L,CE
(5.9)
b-b L ,HE

The cold and hot end leakage parameters may be represented by these leakage

ratio parameters

TnL .
b - ,eel (R 1) (5.10)
L,CE- me. CE+
gl2

771
L } .
HE + 1)
D HE= e' leI (R (5.11)
L, In·
gl2

The preceding definitions of the aIr leakage coefficients lllay be used to

relate the internal and external mass flow rates. By balancing the lllass flow rates

at the gas-in section of the air preheater, we obtain

54
 1 e
m 9 2=m gl°2+ m L ,eel°2+ m L , h el'2 (5.12)

From eqns. (5.6) and (5.7), mL ,eel'2 and m L , h el°2 are as follows

(5.13)

5
L,HE2 e
m - m (5.14)
L,hei2 - (1 + R H E2)
0')
gl_

Substituting the quantities m L ,eel_

0') and m L , h el°2 into eqn. (5.8), we get the

following relation between the internal and external mass flow rates at the gas-in

section of the air preheater.

i -(1 5 L,CE2 + 5 L,HE2) e (5.15 )

m g2- +l+R l+R m gi2
CE2 HE2

The following relations are obtained in the same way.

i ( b L,CE1 5 L,H E1) e

m gl = l+b L2+1+R +1+R m gi2 (5.16)
CE1 HE1

1 e, e (5.17)
a1=
r71
r71 r71 ar°1- (lL , CE1 "')
gl_

(5.18)

The relations between the inlet and outlet flow rates are

(5.19)

55
 ( 5.20)

where £> L ,the air leakage coefficient of the overall air preheater, is the sum of the

air leakage coefficients of both passes.

Figures 5.1 and 5.2 illustrate the air leakage model, internal and external

temperatures (Fig. 5.2) and mass flow rates (Fig. 5.1). In those Figures,

superscript ' i ' and' e ' are used to denote internal and external quantities,

respecti vely.

GAS IN GAS OUT

e e

I ....
1I
m gl2
II
• m g01
I
mL,hel2 mL,cel2 mL,he01 mL,ce01
mg2

I e
ma02 ma2 ma1 mai1
PASS 11 4-mapas- PASS I
AIR AIR
afT IN

mL,he02 mL,ce02 mL,hei1 I mL,cei1

mg1
I J I
mgpass
:1-

Figure 5.1 Internal and external mass flow rates.

56
 GAS IN GAS OUT
I e • e
o T gl2 o T g01
I 1
o T gl2 o T g01

T a02 T al1
T apass
I
4--0 PASS II 0--0 PASS I 0--
AIR AIR
OUT IN

1 1
0 1 gl1
~ T g02
I
T gpass
I

Figure 5.2 Internal and external fluid temperatures.

Internal and external fluid temperature relations can be derived by using

conservation of energy and mass. If cp is constant and (cp)a=(cp)g, then these

relations are simplified to

T a1'1 =T 1
a1'1 = TO a1'1
(5.21)

T a02 -T 1 e (5.22)
- a02 -T
- a02

57
 (5.23)

where

(5.24)

Tgpass= (1+8 L2) (5.25 )

8L,H E1 T + 8 L, CE1 T .
. (1+8L2)Tgpass+(l+RHE1) apass (1+R ) all
Tz CE1 (5.26)
-
gzl- 71

where

(5.27)

T€gol= (l+<5 ) (5.28)

L

58
5.2 Air Preheater Performance With Leakage

The heat transfer effectiveness of the air preheater under consideration is

defined as the ratio of the actual heat transfer rate to the thermodynamically

maximum possible heat transfer, where the air leakage effect is not considered. If

we let cp be constant throughout the air preheater core for both of the fluids and

let the minumum capacity rate be the air capacity rate, then the effectiveness

relations are

7TI e "')
" )_-2::. 0
( T a02 - T all me" effect
[ _ all (5.29)
a- (T e gl_
"')- T al"I)

where

oeffect ={ 0 L,CE1 (T a02 - T ai1 )+(0 L,H E1 +0 L,CE2)( T a02 - T apass)} (5.30)

i i i i
0"( T gi2- T g02)+71( T gi1- T gol)
[g= e (5.31)
"')- T al"I)
C R ( T gl_

where C R is the capacity rate ratio as

( 5.3::!)

Similar hcat transfcr effcctivcncss cxpresslOns may bc dcrivcd for both

passcs of thc air prehcater by using the samc cffcctivcnes: dcfinition.

59
 The air leakage model was added to the computer program which

computes the temperature fields within the air preheater. Using the design data

of the air preheater (see Appendix A), a number of runs were performed with

different air leakage coefficients. At the same time, all other leakage coefficients
>
are kept constant at the following values:

The overall aIr preheater performance decreases with increasing air

leakage. This is shown in Figure 5.3 where the effectiveness of both passes are

also plotted. Figures 5.4 and 5.5 show the effect of air leakage on outlet gas and

air temperatures and mass flow rates, respectively. It is seen that both outlet gas

and air temperatures gradually increase as the air leakage increases.

\Vhen aIr leakage is introduced into the present analysis, it becomes

difficult to identify the reasons for the changes in temperatures of aIr, gas, and

tube metal due to air leakage. because there are many parameters involved with

the heat and mass transfer mechanisms. To illustrate these trends. a table

containing t hose parameters IS prepared for two different cases. one with 0%

leakage. and the other with 10% leakage (see Table 5.1).

As shown in Table 5.1. the average heat transfer coefficient and total heat

60
transfer rate of Pass I are gradually increased as the aIr leakage increases. This

occurs because the gas mass flow rate in Pass I increases by 7.5% while the aIr

mass flow rate of Pass I decreases by 3.5% due to air leakage. In contrast, the

average heat transfer coefficient and total heat transfer rate of Pass II are

decreased with increasing air leakage due to changes in mass flow rates.

When there is 10% total air leakage, Pass II operates with 9.36% lower air

mass flow rate, 0.56% lower average heat transfer coefficient, 2.5% higher gas

mass flow rate, 2.24% higher inlet air temperature and 0.47% lower inlet gas

temperature compared to the no-leakage case. Since less heat is transfered from

the gas to air, the gas leaves Pass II with a higher temperature and mass. On the

other hand, temperatures of air and tube metal within Pass II are increased

because less air mass flows through Pass II, as shown in Figures 5.4 and 5.5. The

gas temperatures decrease close to the gas-inlet section of Pass II due to the cold

and hot end air leakage flows through the gas-inlet section (see Figure 5.4). The

internal gas inlet temperature (r l gil) and internal gas mass flow rate of Pass I

increase with inreasing air leakage (see Table 5.1). The increases in the internal

gas mass flow rate and internal gas inlet temperature of Pass I due to air leakage

cause to increase air, gas, and tube metal temperatures in Pass J (see Figures 5.'1

and 5.6).

61
Table 5.1 Results Qf air leakage analysis.

fJ £=10% increase, %

u1 ' W/m 2 K 29.1625 29.6091 1.53

2 35.0629 34.8659 -0.56
U2 ' W/m K

Q1' kW 75,986.6653 78,564.2883 3.39

Q2' kW 62,192.3926 55,557.3250 -10.67

m 1 aI' kg/ s 536.000 517.195 -3.51

1
m a2' kg/s 536.000 485.853 -9.36

m 1 gl' kg/s 626.840 673.853 7.50

m 1 g2' kg/ s 626.840 642.511 2.50

T Zgil' K 520.103 524.794 0.90

T 1 gal' K 392.052 400.676 2.20

TZC), K 614.820 611.946 -0.4 7

gl-

T Zgo_C). !\' 520.103 529.427 1. 79

T ail' !\' 302.600 302.600 0.00

Tapass, K 442.801 452.729 2.24

T ao_C), I( 555.196 563.322 1.46

 It should be kept in mind that all preceding discussions are based on the

given constant leakage ratio parameters. For' different values of leakage ratio

parameters (RoI' R o2 ' R CEl ' R HE1 , R CE2 ' R HE2 ), the results of the air

leakage analysis may change. It is, therefore, very important to know these

parameters.

One of the most exasperajg problems III air preheater maintenance is the

detection, location, and correction of small leaks. Large leaks can be easily

located, but small leaks are usually hard to detect and still harder to locate

precisely so that they can be repaired. Any leakage model, therefore, will not

represent the real effect of air leakages on the performance of the air preheater.

We were not able to verify the model with the field data. However, we believe

that the presented air leakage model will give approximate information on how

the air leakages affect the performance of the air preheater.

63
 Effectivenesses of the Air Preheater
and its Passes vs. Total Air Leakage
90

• Overall
85 -
•
0
Pass I
Pass II

-- 80 - - --
::.e
0

75 -
- --.
lh
lh
CD
C
CD
70 -
-
>

-
u
CD
65 ~;T
-....
--
-....
--
-.... "t.J
W

60 -
-- --
55 -
.
50 I I I

0 5 10 15 20 25
Total Air Leakage (%)

Figure 5.3 The air leakage effect on the air preheater performance.
 800 I
• •
626.8 kg/s
750 f-mg,in =
f--ma,in = 536.0 kg/s
700 r-- Ta,in = 302.6 K

650

-
~
600 • • • •
c c &;;J LI

550
Q)

-
~

:::J 500
cu
~
Q)
Co 450
E
Q)
I-
400 • - •
350

300 •c Tg,out (K)

Ta,out (K)
250
• Tg,in (K)
200 • • I

o 5 10 15 20 25

air leakage (%)

Figure 5.4 The air leakage effect on internal gas inlet (T 1 0"')'
gl-
external gas outlet (ye gol)' and external air outlet

(T ao_
..,) temperatures.

65
 Outlet Mass Flow Rates of Gas and Air vs.
Total Air Leakage of the Air Preheater

1000
• mg,out (kg/s)
900 l:1 ma,out (kg/s)

---
f/)

C>
800

-
.:lI::

-CI)

co
a:
700

600
~
0
u-
f/)
500
f/)
co
~
400

300

200
a 5 10 15 20 25

Total Air Leakage (%)

Figure 5.5 The air leakage effect on mass flow rates of the air

preheater.

66
 Metal Temperature Variation with leakage
at a Point within Pass I and Pass II
600 I I I I
r-
r-
r-
550 r- 0
• Pass I
Pass II

-u
-
~ 500
I
M n

en
(1)

-
L.
::;:, 450
cL.o
(1)
c.
E 400
-
(1)

-
co
(1)
~
350
• • --
300

I I I I
250
o 5 10 15 20 25
air leakage (%)

Figure 5.6 l\letal temperature variation with leakage at x=.24 m and

y=9.21 m within Pass I and at r=.31 m and y=.48 m within

Pass II.

67
 6. COMPUTER ALGORITHM

6.1 Introduction

A large number of tubular air preheaters in coal fired power plants have

experienced difficulty with plugging and corrosIOn. The presen·ce of sulfur trioxide

in a flue gas elevates the dew point of the gases. Water vapor condences and

forms an acid with the sulfur trioxide and the corrosion of the tubes results in iron

sulphate. To protect the tubes from acid deposition and corrosion, we need to

keep the metal temperature field of the air preheater above the dew point

temperature. This requires that information on the metal temperature field

within the air preheater be available as a function of operating parameters.

Information on metal temperature can be obtained either by direct

measurements or by numerical simulation. Measurements of metal temperatures

are not practical except for research purposes; therefore, numerical simulation of

the heat transfer process is the only alternative.

A computer program' named TPHMT (Tubular Air Preheater Metal

Temperature), developed for a two-pass counter-crossflow tubular air preheater,

computes the metal temperature field of the air preheater as a function of

operating parameters. The mathematical model, presented in Chapter 4.1, IIses

an iterative procedure to obtain the solutions.

68
6.2 General Approach in the TPHMT Code

The computer code TPHMT is made up of the main program and 21

subroutines. A flow chart of the TPHMT code is shown in Figure 6.1.

The code starts by reading the input data and then performs the initial

operations including: calculation of internal and external flow rates of gas and air

through the air preheater and calculation of dimensional velocity and temperature

profiles of gas and air at the inlet of the air preheater.

With the aid of a do-loop, the code performs a cyclical iteration until

convergence IS achieved. The iteration starts with an assumed inlet gas

temperature at Pass I. (see Figure 6.2). Using subroutine APHTM which

performs all necessary mathematical model calculations of a single-pass crossflow

exchanger (Chapter 4.1); the code first calculates the temperature fields of Pass I

starting from A and ending at C, then the same calculations are done for Pass II.

This is the first iteration in the code and more reiterations come after that unless

the calculated inlet gas temperature of Pass I agrees with the assumed one within

an acceptable error. \\Then the convergence criterion is satisfied, the code

computes the heat transfer performance of each pass and the overall air preheater

performance. Finally, it displays the results with the input data in tabular forms.

Results of metal temperature distribution \,,,'ithin the air preheater are stored in a

result file for plotting.

In all cases, con ....ergence to a final steady state solution has been rapid and

generally occurs in about 10 iterations. (This correspondes to a bOli t 3 seconds of

CPU time on a CDC 855 Cyber System.)

69
 INITIAL GUESS FOR INLET
GAS TEMPERATURE OF PASS I

CALCULATE TEMPERATURE
FIELDS OF PASS I

CALCULATE TEMPERATURE
FIELDS OF PASS II

UPDATE INLET GAS TEMPERATURE

OF PASS I WITH THE CALCULATED ONE

NO

YES

CALCULATE HEAT TRANSFER PERFORMANCE

OF THE AIR PREHEATER

PRINT INPUT DATA AND RESULTS

Figure 6.1 Flow chart of computer code TPHMT.

70
 x
GAS IN <-- GAS OUT
II I /\
\/ V y II
---------------
---------------
<-- I I I
<-- C B A
<-- I
<--
<---- PASS II <-- PASS I <----
AIR <-- AIR
OUT <-- IN
<__ Ay
<-- X I
=============== ===============<--
II I===I /\
\/ II

Figure 6.2 The sketch of the tubular air preheater.

6.2.1 Input Data

To run the TPHMT code, the user must supply the following data:

- flow rates of gas and air

_ average gas and air temperatures at the inlet of the air preheater

- air leakage coefficients (if any)

- geometry of the air preheater

- dimensions of the grids

_ inlet gas velocity and temperature variations in nond1771cnsional form

71
 - thermal conductivity of the tube metal

- gas side and air side average pressures

6.2.2 Possible Applications of the Computer Program

The TPHMT code is a computational tool' which enables the user to

analyze the effect of changes in operating parameters and design on the

performance of two-pass counter-crossflow tubular air preheaters and on the

temperature fields of working fluids and tube metal.

Some of the possible applications include:

1. Analysis of the effects of changes in the inlet air temperature

2. Analysis of the effects of air leakage.

3. Analysis of the effects of design changes

4. Analysis of the effects of flow and temperature stratifications

Concerning the first possible application, inlet air temperature control is

typically don;' by the use of steam coils or by recirculating a portion of the hot air

from the air preheater outlet back to the fbrced draft fan inlet.

For the second possible application, as indicated earlier in Chapter 5.1, the

air leakages increases the longer air preheater stays in service. The effects of air

leakage on the performance of the air preheater under consideration and on the

metal temperature distribution within the tube bundle can be easily analyzed by

the TPHMT Code.

As for the third possible application. the effects of a.ny proposed design

change in the su bject air preheater geometry. such as addition or removal of heat

transfer surfaces or a change in its design. can be analyzed by using the TPH~IT
Code before actual modifications are made.

Analysis of other parameters, such as stratification in the inlet air/gas

velocities and temperatures or variations in unit load, is also possible with the

code.

6.3 Results of Sample Calculations

The results presented here were obtained by the TPHMT for a two-pass

counter-crossflow tubular air preheater at Allegheny Power Company's Fort

Martin Power Plant. This aIr preheater has three different tube bundle

arrangements within Pass I as shown in Figure 6.2. The geometry characteristics

and operating conditions of the air preheater are summarized in Appendix A. In

....

these calculations, outlet gas and aIr temperatures and the heat transfer

performance of the aIr preheater as well as the metal temperature distribution

were predicted as a function of the following parameters:

• inlet air temperature

• inlet gas temperature

• mass flow rate of gas

• mass flow rate of air

Uniform temperature and velocity profiles at the aIr preheater inlet were assumed

111 the calculations.

6.3.1 Effect of Inlet A ir Temperature

The TPH!\1T Code was run with inlet air temperature varying from 297 to

308.2 K. The effectiveness of the air preheater remains almost constant as shown

73
in Figure 6.3. The temperatures of the gas and air at the outlet of the air

preheater all increase as the inlet air temperature is increased, as shown in Figures

6.4 and 6.5. The metal temperatures within Pass I decrease sharply as the in~et

air temperature decreases (see Figure 6.6), whereas in Pass II, the effect of inlet

air temperature is small, as shown in Figure 6.7.

6.3.2 Effect of Inlet Gas Temperature

The effectiveness of the air preheater increases gradually as the inlet gas

temperature is increased (see Figure 6.3). The temperature of the air and gas at

the outlet of the air preheater all increase with increasing the inlet air

temperature, as shown in Figures 6.4 and 6.5. The metal temperatures within

Pass II increases sharply as the inlet gas temperature increases (see Figure 6.7),

while the metal temperatures within Pass I increases gradually (see Figure 6.6).

6.3.3 Effect of Gas Flow Rate

Solutions were obtained with the mass flow rate of gas varying from 550

to 690 kg/so As the mass flow rate of gas is increased, the outlet air and gas

temperatures increase as well (see Figures 6.8 and 6.9). The effectiveness and

metal temperatures of the air preheater also increase when the mass flow rate of

gas is increased (see Figures 6.10 ,6.11, and 6.12)

6.3.4 Effect of Air Flow Rate

The mass flow rate of air was varied from 482.4 to 589.6 kg/so It was seen

that the effectiveness of the air preheater (Figure 6.10) and both outlet gas and

air temperatures decrease as the mass flow rate of air is increased as shown in

74
Figure 6.8 and 6.9. The metal temperatures within Pass I and Pass II increase as

the mass flow rate of air is increased, as shown in Figures 6.11 and 6.12.

The metal temperature distributions within Pass I and Pass II are

plotted in Figures 6.13 and 6.14, respectively. These metal temperature

distributions are the results of the code when the design operating conditions of

the air preheater (Appendix A) are used as input data.

75
 TWO-PASS CROSSFLOW AIR PREHEATER

82.0 . I I

D Ta,in = 297.04 K
~ 0 Ta,in = 302.60 K
• Ta,in = 308.16 K

- 81.9 (!)

-
0~

en
en
Q)
c
Q)

--
>
u
81.8 (!)

- _
Q)

w ........." ..... ~_ ....................... __....

...
ca
Q)
81.7 (!)
>
0

I I I
81.6
090 600 610 620 630 640

Inlet Gas Temperature (K)

Figure 6.3 The effect of the inlet gas and air temperature on the

overall effectiveness of the air preheater.

76
 TWO-PASS CROSSFLOW AIR PREHEATER

Outlet air temp. vs. Inlet gas temp.

570
mg,in = 626.84 kg/s
rna, in = 536.00 kg/s
No Air Leakage
-
~
560
...
(1)

-
::J

...
aJ
(1)
Q.

-...
E
(1)

aJ
550

-
-
(1)

::J
0
540
0 Ta,in = 297.04 K
• Ta,in = 302.60 K
c Ta,in = 308.16 K
530
590 600 610 620 630 640

inlet gas temperature (K)

Figurc 6.4 Thc cffcct of thc inlct gas and air tcmpcraturc on thc

outlct air tcmpcraturc of thc air prchcatcr.

I I
 TWO-PASS CROSSFLOW AIR PREHEATER

Outlet gas temp. vs. Inlet gas temp.

420 . . . , - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
rng,in = 626.84 kg/s
rna,in = 536.00 kg/s
415 No Air Leakage
-
~

Q)
410
-
~

::s
C'Q
~
Q)
Co
E 405
-
Q)

(/)
C'Q
Cl

--
400
Q)

::s
0
395 •0
Ta,in = 297.04 K
Ta,in = 302.60 K

• Ta,in = 308.16 K

390 -+--y----r-...---.-..,....-r---r--...---r-....,.----,r----r-...---,.---\
590 600 610 620 ·630 640

inlet gas temperature (K)

Figure 6.5 The effect of the inlet gas and air temperature on the

outlet gas temperature of the air preheater.

78
 TWO-PASS CROSSFLOW AIR PREHEATER

360
mg,in = 626.8 kg/s
S2'
--- 355
ma,in = 536.0 kg/s
en No Air Leakage
en
co
a..
c 350
:E
~
-
c
·0
a..
co
345

• • ----
340
'0
....
CD
::J
ro.... 335
CD
0..
E 330
CD
f-
co
Qi
•
0
Ta,in = 297.04 K
Ta,in = 302.60 K
~ 325
• Ta,in = 308.16 K

320
090 600 610 620 630 640

Inlet Gas Temperature (K)

Figure 6.6 Metal temperature variation with the inlet gas and air

temperatures of the air preheater at x=.24 m and

y=9.21 m within Pass I.

70
 TWO-PASS CROSSFLOW AIR PREHEATER

530
mg,in = 626.8 kg/s
-SZ 525 rna,in = 536.0 kg/s
en
en
ro
No Air Leakage
0-
c: 520
£
~
-
c:
'0
515

0-
ro 510
'0
....
Q)
:::J
505
ro....
Q)
c.
E 500
Q)
t- 0 Ta,in = 297.04 K
~
Q)
~ 495 •
0
Ta,in = 302.60 K
Ta,in = 308.16 K

490
::>90 600 610 620 630 640

Inlet Gas Temperature (K)

Figure 6.7 l\letal temperature variation with the inlet ga.s and air

temperatures of the air preheater at =.31 m and

11=.48 m wit.hin Pass II.

80
 TWO-PASS CROSSFLOW AIR PREHEATER

Outlet gas temp. vs. Gas mass flow rate

460
c- Tg,in = 614.82 K
I - Ta,in = 302.60 K
440 -I-No Air Leakage
-
~
.--€>
...
CIJ

- 420
.--er- ----- ---
----- ------ -----
:::J
...eu
CIJ
c.

-
E
CIJ

Ul
eu
400
.. ---------
~

~
~
~

--
C)
380
CIJ
:::J
0 -
0 ma,in = 482.4 kg/s -
360
•
0
ma,in = 536.0 kg/s
ma,in = 589.6 kg/s
-
-
340 I I I

530 580 630 680 730

gas mass flow rate (kg/s)

Figure 6.8 The effect of the mass flow rates of gas and air on the

outlet gas temperature of the air preheater.

81
 TWO-PASS CROSSFLOW AIR PREHEATER

Outlet air temp. vs. Gas mass flow rate

600
Tg,in = 614.82 K
590 Ta,in = 302.60 K

No Air Leakage
-
~
580

570
...
Q)

-...
::::J
~
Q)
c.
560

E 550
-...
Q)

540
~

--
Q)

::::J
530
• ma, in = 482.4 kg/s
0
520
o ma,in = 536.0 kg/s

510 • ma,in = 589.6 kg/s

500
t>30 580 630 680 730

Gas mass flow rate (kg/s)

Figure 6.9 The effect of the mass flow rates of ga.s a.nd air on the

outlet air temperature of the air preheater.

 TWO-PASS CROSSFLOW AIR PREHEATER

100 I I . I

I--Tg,in = 614.82 K
I- .
95 ~ T a,m = 302.60 K -
--
0~
90
•
en
en
Q)
85 • 0 -
c
Q)
>
- 80 - • rl
- -
--
Co)
Q)

w
-
fl
•
-
...
tU
75
-
-- -
Q)
>
0
70 f-
•
0
rna,in = 482.4 kgls
rna, in = 536.0 kgls - -
65 f-

I
• rna,in = 589.6 kgls

I
-
-
I
60
030 580 630 680 730

Gas Mass Flow Rate (kg/s)

Figure 6.10 The effect of the inlet gas and air mass flow rates on the

overall effectiveness of the air preheater.

83
 TWO-PASS CROSSFLOW AIR PREHEATER

-
~

Gas Mass Flow Rate (kg/s)

Figure 6.11 Metal temperature variation with the inlet gas and air

ma.ss flow rates of the air preheater at x=.24 m and

y=9.21 In within Pass I.

Sol
 TWO-PASS CROSSFLOW AIR PREHEATER

Tg,in = 614.82 K
Ta.in = 302.60 K
en
en
530 No Air Leakage
ctl
a...
c::
£
~ 520
E
'0
a...
ctl 510
o
....
Q)

:::J

'§
Q)
500
c.
E
Q)
I- o rna,in = 482.4 kg/s
ctl 490 • rna,in = 536.0 kg/s
(j)
~ o rna,in = 589.6 kg/s

480 l...-.L-........--'-........--L-..JL..-.&...-...a.-........--1----L........._.&...-.&-...L--.L...........---L.-..J........."J
030 580 630 680 730

Gas Mass Flow Rate (kg/s)

Figure 6.12 Metal temperature variation with the inlet gas and air

mass flow rates of the air preheater at =.31 m and

y=.48 m within Pass II.

85
 METAL TEMPERATURE FIELD Of PASS I, (K)
N
C'l
',', ,-, , , I

,, ,,
'\ '\ I '\
'\ '\ '\

....
'\ '\ '\
,
,,,
'\ '\ t '\
'\ '\ I '\
N
" \ I "
(XJ
,
,,
'\ '\ I '\

'\ '\ \ : \
, ... ...
'\ '\ '\ I '\
,
,,,
'\ '\ '\ I
... ...
"
'\ '\ '\ I
"-
... ...
,"
C'; '\" \ '\
"- '\ '\ I , ...
'\ '\ '\
.,,
"
'\
\.
'\
\
'\ r
I
,
..,
0
'\
'\
'\
'\
'\
'\
'\
'\
'\ , I
I
,,
'\ '\ '\ I , I
CD
"'. . . ~ " t{k" " I
,,,
" '\ 0 \, '\ 0 "' \ ," / , I

..,.., ,,
f
'\ '\ '\ I I

" " \ J : I
vi " '\ ,,: I
\
\~,
I
\, " '\ , r
,,
... ,-,
E ,, \
I

,, ,
~
\
\
I ... ...
r ...
..,
CD
\
\
...
\
\
\ I
I ,
I ...
.... "
'\
'\
" I
I
, I
I

\ \ I I
, , I
I
I
" ' , , ' I
..,
C'l ... \. ",',' I
I

.., ......
... ... ',,\ ',,',' I
I
I

... ... " '\ I I

, I
... , " ' ... / /
... ... , I
, ... \
, I ,-,
N
.... ... ...
\
I

N ... ... " \ II I

I

,
\
\
, I
, I

, I
I

... ,
,, ,,
\ I I
\ I

....
lf1
... ...
, I
,
,,
, I
\ ~ I
~- -
....
(XJ
,
o +---.;:;,,"'l""----'~':.....---~---;;....,..- ...--"T""-....- . . . ':"""---__:-.. . .- '--_:_---_l
0.31 0.93 1. 55 2.17 -:.03 1.65 5.27 5.89
x
CAS IN <-- GAS Ot..'"":
II I /\
\1 V Y II
----------- <---------------
111111111
<--

<--
bill f I 1
<---- PASS II <-- PASS
AIR <--
OUT <--
<-- y
111111111 <-- 11111 x 1
------------
II I---I 1\
\1 11

Figure 6.13 ~Ietal temperature distribution within Pass J for baseline

conditions.

:'6
 METAL TEMPERATURE FIELD Of PASS II, (K)
N
oi

....
N

-, -
<Xl
,

"-
N
~COa
"-

Cl
n
to

n
n
Ln

E
r lO
n
....

CJ'
n
n

N
....
N

....
lf1
~
<.0
o
....
<Xl

Cl
I

0.31 0.93 1. SS 2.17 2.79 3.~1 ~.03

\~.6S 5.27 5.29
X(mJ
GAS IN C,;S ct."'":'
II I /\
\/ V IIY
------------ --------------
<-- I
II1IIII1I bill r
<--

<--
I
A

<---- PASS II <-- ?;Ss <----

AIR <-- A:P.
0 ..." 7 :::::~;

<--
II
}'
1111I1111 <-- III I X

-----------
II
---------------<--
I---I
/\
\/ II

Figure 6.1~ :'Ictal temperature distribution within Pass II for baseline

conditions.
 7. SUMMARY AND CONCLUSIONS

A mathematical model has been developed to predict the metal

temperature field, both working fluid temperature fields and heat transfer

characteristics of crossflow type tubular air preheaters. A computer program

applying this mathematical model to a two-pass counter-crossflow tubular air

preheater has been created. The effects of air preheater operating parameters on

the tube metal temperature and the performance of this particular air preheater

are easily evaluated with the aid of the computer program. Prediction of the tube

metal temperature field would enable the utility to operate the air preheater at

the onset of acid deposition and corrosion.

Based on the results of this study, we have concluded that the presented

computer modeling technique can be used

To analyse the effects of air leakage on the performance of the

air preheater.

To study acid deposition and corrosIOn phenomena on the tubes.

To modify air preheater geometries in existing units.

- To design new air preheaters for new units.

\Vork is continuing at thc Encrgy Rcscarch Ccntcr to \'crify this computcr

modcling tcchniquc with ficld data and apply it to diffcrcnt crossflow

arrangcmcnts of air prchcatcrs. For example, it is prcsently being used to study

threc and four-pass counter-crossflow air preheaters.

ss
 8. RECOMMENDATIONS FOR FUTURE WORK

Tubular air Preheaters may be arranged with more than two passes. As

long as the passes of the air preheater have a crossflow arrangement, the basic

approach used in the computer program (Chapter 6.) will enable us to find the

temperature fields and heat transfer characteristics of the air preheater. However,

each additional case requires a special version of the code.

Figure 8.1 shows some of these multi-pass crossflow air preheater

arrangements with various directions of the working fluids. As seen from these

arrangements, some of the passes have different flow directions and boundary

conditions. Therefore, the equations and calculation sequence described In

Chapter 4.1 should be modified for those passes. There are four possible gas/air

flow direction com binat.ions. For each gas/air flow direction com binat.ion there

are four boundary conditions (see Fig. 8.2) yielding 16 different cases which need

to be modelled computationally. A separate subroutine should be written for each

of these 16 different cases, if an expansion of the capabilities of the TPH1\IT code

is desired.
 GAS OUT

AIR
82 IN

03

AIR
4--4-- cur 82

GftS IN

a)

r GftS
, r GftS
,

AI R AI R
IN our
- I-e. C2 D2 C4 04 - f-

GftS OUT .......GftS /'" GftS IN

~
b)

Figure 8.1 Multi-pass crossflO'v air preheaters; a) three-pass

COli n ter-crossflow. b) fOll r-pass cou n ter-crossflow

90
 a) Possibla Flow Diractions of Working Fluids

GAS x
+ tY

• •
A) 4 - AIR B) 4 - AIR

.jY
Jsx
x GAS

Yt +
AIR

•
AIR.
C) - . D) - .

Y t.. +
x GAS

b) Possible Known Boundary Conditions

• •
- • •
Represents the known boundary condition at that corner

Figure 8.2 a) Possible now dircctions. and b) possiblc givcn boundary

conditions of a singlc-pass cross now exchanger.

91
 REFERENCES

[ 1 ) Kays, W. and London, A. L., Compact Heat Exchangers, 3rd ed.,

McGraw, New York, 1984.

[ 2 ) Mason, J. L., 'Heat Transfer in Crossflow,' Proc. AppI. Mechanics,

2nd U.S. Nat. Congress, 1954, p. 80l.

[3 ) Chung, H. L., 'A New Simplified Formula for Crossflow Heat Exchanger

Effectiveness,' ASME Journal of Heat Transfer, Vol. 109, 1987,pp. 521-22.

[ 4 ) Bankhe, G. D. and Howard, C. P " 'The Effect of Longitudinal Heat

Conduction on Periodic-Flow Heat Exchanger Performance,' Trans.

ASM E, Journal of Engineering for Power, Vol. 86, Series A, 1964,

pp. 105-120.

[5 ) Schultz, B. H., 'Regenerators with Longitudinal Heat Conduction,'

1M E- ASM E General Discussion on Heat Transfer, London, England,

1951, pp. 440-43.

[6 ] Hahnemann, H. \V., 'Approximate Calculation of Thermal Ratios in Heat

Exchangers Including Heat Conduction in Direction of Flow,' National Gas

Turbine Establishment Memorandum, No. M36, 1948.

[ 7] Kroeger, P. G., 'Performance Deterioration in High Effectiveness Heat

Exchanger Due to Axial Heat Conduction Effects,' 1966 Cryogenic

Engineering Conference, Boulder, Colo.

[S] Landau, H. G. and Hlinka, J. \V., 'Steady State Temperature Distribution

in a Counteri1ow Heat Exchanger Including Longitudinal Conduction in

the Wall,' ASME paper, No. 60- WA-236, 1960.

[9 ) Chiou, J. P _, 'Thermal Performance Deterioration in Crossi1ow Heat

92
 Exchanger Due to Longitudinal Heat Conduction in the Wall,' ASME

76- WAf HT-8, paper present'ed at ASME Winter Annual Meeting, 1976.

[ 10] Shah, R. K. and A. L. London, 'Laminar Flow Forced Convection in

Ducts,' Supplement 1 to Advances in Heat Transfer, Academic Press,

New York, 1978.

[ 11] Shah, R. K. and M. S. Bhatti, in S. Kakac, R. K. Shah, and W. Aung,

Eds., Handbook of Single-Phase Convective Heat Transfer,

Wiley-Interscience, New York, 1987.

[ 12] Heaton, H. S., W. C. Reynolds, and W. M. Kays, 'Heat Transfer in

Annular Passages. Simultaneous Development of Velocity and

Temperature Field in Laminar Flow,' Int. J. Heat A1ass Transfer,

Vol. 7, 1964, pp. 763-781.

[ 13] Kays, W., 'Numerical Solutions for Laminar-Flow Heat Transfer in

Circular Tubes,' Trans. ASME, Vol. 77, 1955, pp. 1265-1274.

[ 14] Petukhov, B. S. and V. N. Poppv, 'Theoretical Calculation of Heat

Exchange and Frictional Resistance in Turbulent Flow in Tubes of an

Incompressible Fluid with Variable Physical Properties,' High Temperature

,Vol. 1, No.1, 1963, pp. 69-83.

[ 15] Petukhov, B. S., in T. F. Irvine and J. P. Hartnett, Eds., Advances in

Heat Transfer, Vol. 6, Academic Press, New York, 1970.

[ 16] Gnielinski, V., 1nt. Chern. Eng., Vol. 16, p. 359, 1976.

[ 17 1 Kays, \\T. and H. C. Perkins, 'Forced Convection, Internal Flow in Ducts.'

Handbook of Heat Transfer, Fundamentals, 2nd ed., ed. \\T. M. Rohsenov,,·.

J. P. Hartnett, and E. N. Ganic, Chapter 7, McGraw, New York, 1985.

[ 18 1 ~li1ls, A. F .. 'Experimental In\"cstigation of Turbulent Heat Transfer in

93
 the Entrance Region of a Circular Conduit,' J. Mech. Eng. Sci., Vol. 4,

1962, pp. 63-77.

[ 19] Nikuradse, J. : Forsch. Arb. Ing. - Wes., no. 361, 1933; English trans!.,

NACA TM 1292.

[ 20] Schlichting, R., Experimentelle Untersuchungen Zum Rauhigkeitsproblem,

lng.-Arch., Vol. 7, pp. 1-34,1936; English trans!. Proc ASME, 1936.

[ 21] Moody, L. F., 'Friction Factors for Pipe Flow,' Trans. ASME, Vol. 66,

1944, pp. 671-684.

[ 22] Norris, R. R., 'Some Simple Approximate Heat Transfer Correlations for

Turbulent Flow in Ducts with Rough Surfaces,' Augmentation of

Convective Heat and Mass Transfer, ASME, New York, Dec. 1970.

[ 23] Zukuaskas, A. A., Heat T7'ansfer in Tube Banks in Crossflow, Hemisphere

Publishing Corp., New York, 1988.

[ 24] Incropera, F. P. and D. P. Dewitt, Fundamentals of Heat Mass Transfer,

3rd ed., John Wiley & Sons, New York, 1990.

[25] Korst, H. H., 'Mean Temperature Difference in Multi-Pass Crossf1ow Heat

Exchangers,' Proceedings of the First U.S. National Congress of Applied

Mechanics, ASME, New York, 1952, pp. 949-55.

[ 26] Stevens, R. A., J. Fernandez, and J. R. \\'oolf, 'Mean-Temperature

Difference in One, Two and Three-Pass Crossf1ow Exchangers,'

Trans. ASME, Vol. 79, 1957, pp. 287-97.

94
 APPENDIX A

An example Output of the TPHMT Code for the Design Operating

Conditions of the Air Preheater

at Allegheny Power Company's Fort Martin Power Plant

95
1I1I1I1 PPPPPP H H M M I I II I I I
T P P H H M M M M T
T PPPPPP HHHHHH M MM M T
T P H H M M T
T P H H M M T

A THO DIMENSIONAL HEAT TRANSFER MODEL FOR

A THO-PASS CROSSFLOH AIR PRE HEATER

VERSION 1.1

by
ALI YILMAZ

x
GAS IN <-- GAS OUT
I1 I /,
,/ V y II
=============== ===============
111111111111111<--111111 1 I 1 I
11111I1111I1I11<--llclll B I A I
11111111111111\<--111\11 I I I I
III III <--III I I
<----1 I I PASS II \1 I <--III PASS I I 1<----
AIR 1 I I I I I <--I I I I I AIR
OUT I I I II I \ I I I \\ 1 I I <--I I I III I I I I IN
\ I III I III 111\\\ <--111\1\ I I I I -y
111111111111111<--111111 I I 1 I x I
=============== ===============<--
1I I===I /, I
V II I
------- 1

ENERGY RESEARCH CENTER

200 PACKARD LABORATORY, BLD. 19
LEHIGH UNIVERSITY
BETHLEHEM, PA 18015
PHONE: (215) 758-4090
 \

INLET GAS VELOCITY PROFILE AT PASS II, m/s vs. x, m

Vg -> 7.97 7.97 7.97 7.97 7.97 7.97 7.97 7.97 7.97 7.97
x -> 5.91 5.29 4.67 4.04 3.42 2.80 2.18 1.56 .93 .31

INLET AIR VELOCITY PROFILE AT PASS I, m/s vs. y, m

Va -> 1. 96 1. 96 1. 96 1.96 1. 96 1. 96 1. 96 1. 96 1.96 1.96
Y -> .48 1.45 2.42 3.39 4.36 5.33 6.30 7.27 8.24 9.21

FLOW RATES AND AVERAGE TEMPERATURES OF GAS AND AIR

AT INLET AND OUTLET SECTIONS OF EACH PASS
APH TOTAL AIR LEAKAGE, % = 0.00

Mg,in = 626.8 kg/s Mg,out 626.8 kg/s

Tg,in 614.8 K Tg,out = 388.3 K

X
GAS IN <-- GAS OUT
II I A
V V y II
=============== ===============
111111111111111<--111111 I I I I
IIIIIIIIIIIIIII<--IICIII B I A I
Ma ,0= 536.0 kg/s II111 11111I11 II <--111111 I I I I Ha, in= 536.0 kg/s
Ta,o= 554.0 K I I I PASS II I I 1<--1 I I PASS I I I Ta,in= 297.0 K
<----111 ------- I I I<--III ------- I 1<----
AIR III E2=65.9 111<--111 El=65.1 I I AIR
OUT II I I I I I I 11I I I I I<--I I II 1 I I I I I IN
I I I I I I I II I I II I I<--II I I I I I I I I ~ y
111111111111111<--111111 I I I I x I
=============== ===============<--
II 1===1 A I
V II I
------- 1
Tg,pass = 518.4 K
E overall= 82.5 %

97
 AIR LEAKAGE ANALYSIS

INTERNAL AND EXTERNAL AVERAGE MASS FLOH RATES AND TEMPERATURES

EXTERNAL GAS INLET FLOH RATE OF PASS II, kg/s = 626.840

INTERNAL GAS FLOH RATE OF PASS II, kg/s = 626.840
PASSAGE GAS FLOH RATE, kg/s = 626.840
INTERNAL GAS FLOH RATE OF PASS I, kg/s 626.840
EXTERNAL GAS OUTLET FLOH RATE OF PASS I, kg/s 626.840

EXTERNAL AIR INLET FLOH RATE OF PASS I, kg/s 536.000

INTERNAL AIR FLOH RATE OF PASS I, kg/s = 536.000
PASSAGE AIR FLOH RATE, kg/s = 536.000
INTERNAL AIR FLOH RATE OF PASS II, kg/s = 536.000
AIR OUTLET FLOH RATE OF PASS II, kg/s = 536.000

EXTERNAL GAS INLET TEMPERATURE OF PASS II, K = 614.820

INTERNAL GAS INLET TE~PERATURE OF PASS II, K = 614.820
INTERNAL GAS OUTLET TEMPERATURE OF PASS II, K = 518.373
PASSAGE GAS TEMPERATURE, K 518.373
INTERNAL GAS INLET TEMPERATURE OF PASS I, K = 518.373
INTERNAL GAS OUTLET TEMPERATURE OF PASS I, K = 388.287
EXTERNAL GAS OUTLET TEMPERATURE OF PASS I, K 388.287

AIR INLET TEMPERATURE OF PASS I, K 297.040

PASSAGE AIR TEMPERATURE, K = 439.571
AIR OUTLET TEMPERATURE OF PASS II, K 554.049

TAPH TOTAL AIR LEAKAGE % = 0.000

TOTAL AIR LEAKAGE OF PASS I, % 0.000
HOT END AIR LEAKAGE OF PASS I, % 0.000
COLD END AIR LEAKAGE OF PASS I, % 0.000
TOTAL AIR LEAKAGE OF PASS II, % = 0.000
HOT END AIR LEAKAGE OF PASS II, % = 0.000
COLD END AIR LEAKAGE OF PASS II, % 0.000
 PASS I

PART "A"

SOME OF THE INLET PARAMETERS USED IN THE TPHMT

CODE, AND THEIR VALUES

**********************************************************************

GAS SIDE AVG. PRESSURE, N/m-2 97591.5000

AIR SIDE AVG. PRESSURE, N/m-2 = 105053.8000
DEPTH OF THE HEAT EXCHANGER lin the z directionl,m 22.8600
HIDTH OF THE HEAT EXCHANGER lin the x direction),m .9525
LENGTH OF THE HEAT EXCHANGERlin the y directionl,m 9.6933
INSIDE DIAMETERS OF TUBES, m = .0416
OUTSIDE DIAMETERS OF TUBES, m .0508
THERMAL CONDUCTIVITY OF TUBE METAL. H/m-K 63.9000
GAS SIDE HEAT TRANSFER AREA PER UNIT CORE VOLUME
m-2/m-3 = 9.0271
TRANSVERSE PITCH OF THE TUBE BANK, m .2432
LONGITUDINAL PITCIi OF THE TUBE BANK. m .0595
INDICATES HOH TUBES ARE ARRANGED IN THE HEAT
EXCHANGER 10-INLINE. I-STAGGERED) o
**********************************************************************

PART "B"

SOME OF THE INLET PARAMETERS USED IN THE TPHMT

CODE, AND THEIR VALUES

**********************************************************************

GAS SIDE AVG. PRESSURE. N/m-2 97591. 5000

AIR SIDE AVG. PRE3SURE, N/m-2 105053.8000
DEPTH OF THE HeAT EXCHANGER lin the z directionl.m = 22.8600
HIDTH OF THE HEAT EY.CHANGER lin the x directionl.m 1.1938
LENGTH OF THE HEAT EXCHANGER I in the y direction l.m 9.6933
INSIDE DIAMETERS OF TUBES. m .0416
OUTSIDE DIAMETERS OF TUBES. m .0508
THERMAL CONDUCTIVITY OF TUBE METAL. H/m-K 63.9000
GAS SIDE HEAT TRANSFER AREA PER UNIT CORE VOLUME
m-2/m-3 13.3129
TRANSVERSE PITCH OF THE TUBE BANK. m .1645
LONGITUDINAL PITCH OF THE TUBE BANK. m .0597
INDICATES HOH TUBES ARE ARRANGED IN THE HEAT
EXCHANGER 10-INLINE. I-STAGGERED) o

99
 PART "C"

SOME OF THE INLET PARAMETERS USED IN THE TPHMT

CODE, AND THEIR VALUES

**********************************************************************

GAS SIDE AVG. PRESSURE, N/m-2 97591.5000

AIR SIDE AVG. PRESSURE, N/m-2 105053.8000
DEPTH OF THE HEAT EXCHANGER lin the z directionl,m = 22.8600
HIDTH OF THE HEAT EXCHANGER lin the x directionl,m 4.7625
LENGTH OF THE HEAT EXCHANGERlin the y directionl,m 9.6933
INSIDE DIAMETERS OF TUBES, m = .0416
OUTSIDE DIAMETERS OF TUBES, m = .0508
THERMAL CONDUCTIVITY OF TUBE METAL, H/m-K 63.9000
GAS SIDE HEAT TRANSFER AREA PER UNIT CORE VOLUME
m-2/m-3 25.7886
TRANSVERSE PITCH OF THE TUBE BANK, m .0819
LONGITUDINAL PITCH OF THE TUBE BANK, m = .0618
INDICATES HOH TUBES ARE ARRANGED IN THE HEAT
EXCHANGER 10-INLINE, I-STAGGERED) o
**********************************************************************

PASS I I

SOME OF THE INLET PARAMETERS USED IN THE TPHMT

CODE, AND THEIR VALUES

*****************************~************************ ****************

GAS SIDE AVG. PRESSURE, N/m-2 = 97591.5000

AIR SIDE AVG. PRESSURE, N/m-2 105053.8000
DEPTH OF THE HEAT EXCHANGER lin the z directionl,m 22.8600
I-UDTH OF THE HEAT EXCHANGER I in t;.,." x direct ion I ,m 6.2230
LENGTH OF THE HEAT EXCH~NG~Rlin th~ y direciionl,m 9.6933
INSIDE DIAMETERS OF TUBES, m .0416
OUTSIDE DIAMETERS OF TUBES, m .0508
THERMAL CONDUCTIVITY OF TUBE METAL, H/m-K 63.9000
GAS SIDE HEAT TRANSFER AREA PER UNIT CORE VOLUME
m-2/m-3 25.8877
TRANSVERSE PITCH OF TH
~ UBE BANK, m
LONGITUDINAL PIrCH OF TH TUBE BANK, m
INDICATES HOW TUBES ARE ARRANGED IN THE HEAT
.0819
.0622
EXCHANGER 10-INLINE, I-STAGGERED) o
**********************************************************************

100
 RESULTS

PART "A"

TOTAL HEAT TRANSFER RATE, IkHJ 5959.1931

CAPACITY RATE OF AIR, IH/KJ = 536341.6277
CAPACITY RATE OF GAS, IH/KJ 37386.6393
CAPACITY RATE RATIO, lnondim. J = .0697
NUMBER OF TRANSFER UNITS lnondim. J = 1.3249
EFFECTIVENESS OF HEAT EXCHANGER = .7202
AVG. OVERALL HEAT TRANSFER COEF. IH/m-2-KJ 25.9969
TOTAL GAS-SIDE HEAT TRANSFER AREAlm-2J = 1905.2965
VOL. OF HEAT EXCHANGER CORE Im-3J 211.0634
*****************************************************************

PART "B"

*****************************************************************
TOTAL HEAT TRANSFER RATE, IkHJ = 10608.4277
CAPACITY RATE OF AIR, IH/KJ = 537420.4742
CAPACITY RATE OF GAS, IH/KJ 69156.0561
CAPACITY RATE RATIO, Inondim. J = .1287
NUMBER OF TRANSFER UNITS lnondim. J 1.4014
EFFECTIVENESS OF HEAT EXCHANGER .7297
AVG. OVERALL HEAT TRANSFER COEF. IH/m-2-KJ 27.5193
TOTAL GAS-SIDE HEAT TRANSFER AREAlm-2J 3521.7034
VOL. OF HEAT EXCHANGER CORE Im-3J = 264.5328
******~*******************************;~*************************

PART "C"

*****************************************************************
TOTAL HEAT TRANSFER RATE, IkHJ 60626.8690
CAPACITY RATE OF AIR, IH/KJ = 543295.1872
CAPACITY RATE OF GAS, IH/KJ 537016.9623
CAPA=ITY RATE RATIO, lnondim. .9884
NUMBeR OF TRANSFER UNITS lnondim. 1.7023
EFFEC fIVENESS OF HEAT EXCHANGER .5927
AVG. OVERALL HEAT TRANSFER CO~F. (H/m-2-KI 33.5903
TOTAL GAS-SIDE HEAT TRANSFER AREAlm-2J 27215.0922
VOL. OF HEAT EXCHANGER CORE Im-3J 1055.3168
*****************************************************************

101
 PASS I

TOTAL HEAT TRANSFER RATE, IkHl 77194.4898

CAPACITY RATE OF AIR, IH/Kl; 539019.0964
CAPACITY RATE OF GAS, IH/Kl 643559.6577
CAPACITY RATE RATIO, lnondim. I .8376
NUMBER OF TRANSFER UNITS lnondim. I ; 1.7583
EFFECTIVENESS OF HEAT EXCHANGER ; .6506
AVG. OVERALL HEAT TRANSFER COEF. (H/m-2-Kl ; 29.0355
TOTAL GAS-SIDE HEAT TRANSFER AREAlm-21 ; 32642.0920
VOL. OF HEAT EXCHANGER CORE Im-31 ; 1530.9130
*****************************************************************

PASS II

*****************************************************************
TOTAL HEAT TRANSFER RATE, IkHl ; 63319.0963
CAPACITY RATE OF AIR, IH/Kl; 554239.3802
CAPACITY RATE OF GAS, IH/Kl 655317.3345
CAPACITY RATE RATIO, lnondim. I ; .8458
NUMBER OF TRANSFER UNITS I nondim. I ; 2.2542
EFFECTIVENESS OF HEAT EXCHANGER ; .6587
AVG. OVERALL HEAT TRANSFER COEF. IH/m-2-Kl ; 34.9984
TOTAL GAS-SIDE HEAT TRANSFER AREAlm-21 35697.7251
VOL. OF HEAT EXCHANGER CORE Im-31 ; 1378.9473
*****************************************************************

AIR PRE HEATER

I overall I

*****************************************************************
TOTAL HEAT TRANSFER RATE, IkHl ; 140513.5861
EFFECTIVENESS OF HEAT EXCHANGER .8248
AVG. OVERALL HEAT TRANSFER COEF. lH/m-2-Kl ; 32.0169
TOTAL GAS-SIDE HEAT TRANSFER AREAlm-21 68339.8172
VOL. OF HEAT EXCHANGER CORE (m-31 2909.8603
*****************************************************************

10~
 2-Dimensional Representation of Pass I
on xy plane
DXl j J
1<------>1 Tg l,j

1 - I 1
I 1 1 1
Tm 1,1 IDYliJ Ta l,jl Tm l,j Ta l,j+l Tm I,N I
I I 1 I 1
1 V 1 I I
----------1I
I
1
1
Tg i,l Tg i,j 1
1
----------1
1 1
1 1 I
Ta i,l Tm i,l Ta i,jl Tm i,j Ta i,j+l 1 Tm i,N 1 <--- AIR
1 1 1
1 1 I
Tg i+l,1 Tg i+l,j
----------11
1
I
I
Tg H,I Tg H,N 1
----------1
1 1
1 1
Ta H,I Tm H,I Ta H,2 Tm H,j Ta H,N 1 Tm H,N 1 Ta H,N+1
1 I
I 1
Tg H+l,1 Tg H+l,N
GAS
A
Y -
1
II x <--
-
i, j 1 2 3 4 5 6 7 8 9 10
Dimensions of the Smaller Heat Exchanger Assemblies, m

OXI j J .79 .79 .79 .79 .79 .79 .60 .60 .48 .48
OYI i J .97 .97 .97 .97 .97 .97 .97 .97 .97 .97

Coordinates of the Smaller Heat Exchanger Asse",bl ies, m

Xl j I 6.51 5.72 4.92 4.13 3.34 2.54 1.85 1. 25 .71 .24

YI i I 9.21 8.24 7.27 6.30 5.33 4.36 3.39 2.42 1.45 .48

103
 TEMPERATURE FIELD OF PASS I :

FLUE GAS TEMPERATURE VARIATION IN THE EXCHANGER: Tgli,jl, K

j 1 2 3 4 5 6 7 8 9 10
i
1 439.3 427.6 414.6 400.2 384.3 367.1 368.8 362.0 361.3 357.6
2 447.6 436.3 423.6 409.3 393.3 375.5 377.3 370.3 369.7 365.8
3 456.0 445.3 433.1 419.1 403.1 385.0 386.9 379.7 379.2 375.2
4 464.5 454.5 443.0 429.5 413.8 395.6 397.5 390.4 389.8 385.9
5 473.0 463.9 453.2 440.5 425.4 407.5 409.4 402.4 402.0 398.0
6 481.4 473.4 463.8 452.2 438.0 420.8 422.8 416.0 415.6 411. 9
7 489.7 483.0 474.7 464.4 451. 7 435.8 437.7 431.5 431.2 427.7
8 497.7 492.4 485.7 477.3 466.5 452.7 454.3 448.9 448.7 445.7
9 505.3 501.6 496.8 490.6 482.5 471.7 473.0 468.7 468.6 466.3
10 512.4 510.4 507.8 504.4 499.6 493.0 493.9 491.3 491.3 489.8
11 518.4 518.4 518.4 518.4 518.4 518.4 518.4 518.4 518.4 518.4

AIR TEMPERATURE VARIATION IN THE EXCHANGER: Tali,jl, K

j 1 2 3 4 5 6 7 8 9 10 11
i
1 399.7 386.2 371. 9 357.1 342.1 327.4 313.5 308.1 302.8 299.9 297.0
2 407.4 393.6 378.9 363.3 347.3 331. 2 315.6 309.6 303.5 300.3 297.0
3 415.5 401.6 386.4 370.2 353.1 335.5 318.1 311. 2 304.4 300.7 297.0
4 424.0 410.1 394.6 377.7 359.6 340.4 320.8 313.2 305.5 301.3 297.0
5 433.0 419.1 403.5 386.0 366.8 346.0 323.9 315.4 306.6 301. 9 297.0
6 442.3 428.7 413.0 395.1 374.9 352.3 327.5 317.9 307.9 302.5 297.0
7 452.0 438.8 423.3 405.1 383.9 359.4 331.5 320.7 309.5 303.3 297.0
8 462.0 449.5 434.4 416.1 394.0 367.6 336.1 324.1 311.2 304.3 297.0
9 472.6 461. 0 446.5 428.3 405.6 377.1 341.5 328.0 313.3 305.3 297.0
10 487.2 477.4 464.3 447.0 423.9 392.7 350.3 334.4 316.7 307.1 297.0

METAL TEMPERATURE VARIATION IN THE EXCHANGER: Tml i,j), K

j 1 2 3 4 5 6 7 8 9 10
i

1 417.3 404.4 390.4 375.5 359.8 343.7 345.2 338.9 338.1 334.7
2 425.3 412.4 398.3 383.0 366.7 349.5 351.2 344.4 343.6 339.9
3 433.6 42(1.9 406.8 391. 2 374.3 356.1 357.8 350.6 349.9 345.9
4 442.2 429.8 415.8 400.1 382.6 363.4 365.3 357.7 357.0 352.8
5 451.0 439.1 425.4 409.7 391.8 371. 7 373.8 365.8 365.1 360.8
6 460.0 448.7 435.5 420.0 402.0 381.1 383.3 375.1 374.4 369.9
7 469.1 458.7 446.2 431. 2 413.2 391. 7 394.1 385.6 385.1 380.4
8 478.3 468.9 457.4 443.1 425.6 403.8 406.3 397.7 397.3 392.5
9 487.6 479.5 469.3 456.2 439.4 417.8 420.4 411.9 411.6 406.9
10 499.9 494.0 486.1 475.4 460.8 440.6 443.2 435.3 435.2 430.8

10-1
 2-Dimensional Representation of Pass II
on xy plane

GAS X <--
1 I
V y V
DXI j)
1<------>1 Tg l,j

I - 1
1 1 1
Tm 1,1 IDYlil Ta l,jl Tm l,j Ta l,j+l Tm l,N
I 1 I
I V 1

Tg i,l Tg i,j
--------~~-------I-
1 I I
Ta i,l Tm i,l 1 Ta i,jl Tm i,j I Ta i,j+l Tm i,N <--- AIR
1 1 1
I 1 1
Tg i+l,l Tg i+l,j

Tg H,l Tg H,N
----------1
1 I
I I
Ta H,l Tm H,l Ta H,2 Tm H,j Ta H,N I Tm H,N I Ta H,N+l
I I
I 1
Tg H+l,l Tg H+l,N

i, j 1 2 3 4 5 6 7 8 9 10

Dimensions of the Smaller Heat Exchanger Assemblies, m

DXl j I .62 .62 .62 .62 .62 .62 .62 .62 .62 .62
DYl i I .97 .97 .97 .97 .97 .97 .97 .97 .97 .97

Coordinates of the Smaller Heat Exchanger Assemblies, m

Xl j) 5.91 5.29 4.67 4.04 3.42 2.80 2.18 1.56 .93 .31
YI i I .48 1.45 2.42 3.39 4.36 5.33 6.30 7.27 8.24 9.21

105
 TEMPERATURE FIELD OF PASS II :
FLUE GAS TEMPERATURE VARIATION IN THE EXCHANGER: Tgl i,j I, K

j 1 2 3 4 5 6 7 8 9 10
i
1 614.8 614.8 614.8 614.8 614.8 614.8 614.8 614.8 614.8 614.8
2 609.9 608.7 607.3 605.5 603.2 600.4 596.8 592.2 586.3 578.7
3 603.9 601. 7 598.9 595.5 591.5 586.5 580.3 572.8 563.4 551. 7
4 597.5 594.2 590.3 585.6 580.0 573.4 565.4 555.9 544.3 530.4
5 590.7 586.5 581.6 575.9 569.2 561.3 552.1 541.3 528.7 513.7
6 583.8 578.9 573.2 566.7 559.1 550.4 540.5 529.1 516.0 501. 0
7 577.0 571.4 565.1 558.0 549.9 540.8 530.5 519.0 506.2 491.8
8 570.2 564.2 557.5 550.0 541.6 532.4 522.2 511.0 498.8 485.6
9 563.8 557.5 550.5 54"'2.8 534.4 525.3 515.5 504.9 493.7 481. 9
10 557.7 551. 2 544.2 536.5 528.3 519.6 510.3 500.7 490.6 480.4
11 552.2 545.7 538.7 531.3 523.5 515.4 507.0 498.5 489.9 481.5

AIR TEMPERATURE VARIATION IN THE EXCHANGER: Tali,jl, K

j 1 2 3 4 5 6 7 8 9 10 11
i
1 590.3 584.5 577.3 568.5 557.5 543.7 526.6 505.1 478.0 443.6 399.7
2 577.8 570.8 562.4 552.5 540.7 526.8 510.2 490.6 467.4 439.9 407.4
3 569.3 561. 7 552.9 542.7 530.9 517.4 501.9 484.1 463.9 441.1 415.5
4 561.4 553.5 544.4 534.2 522.8 509.9 495.6 479.9 462.6 443.9 424.0
5 554.2 546.1 537.0 527.1 516.1 504.2 491.3 477.5 462.9 448.0 433.0
6 547.6 539.5 530.7 521.1 510.8 499.9 488.5 476.7 464.8 453.2 442.3
7 541. 7 533.8 525.3 516.3 506.9 497.2 487.3 477.5 468.0 459.3 452.0
8 536.6 529.0 521.0 512.7 504.3 495.8 487.5 479.5 472.4 466.4 462.0
9 532.3 525.1 517.8 510.3 502.9 495.7 489.0 482.9 477.9 474.3 472.6
10 529.3 522.8 516.3 509.9 503.8 498.2 493.3 489.4 486.8 485.9 487.2

METAL TEMPERATURE VARIATION IN THE EXCHANGER: Tmli,jl, K

j 1 2 3 4 5 6 7 8 9 10

1 600.4 596.9 592.7 587.3 580.7 572.3 561. 7 548.3 531.0 508.4
2 589.8 584.9 579.0 572.0 563.6 553.5 541.3 526.7 509.0 487.4
3 582.0 576.2 569.5 561.7 552.5 541.8 529.4 514.9 497.9 478.2
4 574.4 568.0 560.7 552.4 542.8 532.0 519.6 505.7 490.0 472.5
5 567.1 560.2 552.6 544.0 534.4 523.7 511.8 498.8 484.6 469.4
6 560.1 553.1 545.2 536.6 527.1 516.9 505.8 494.0 481.5 468.6
7 553.7 546.5 538.7 530.3 521. 2 511.6 501.4 490.9 480.3 469.7
8 547.9 540.7 533.1 525.0 516.5 507.7 498.7 489.6 480.8 472.4
9 542.7 535.7 528.4 520.8 513 .1 505.2 497.4 489.9 482.8 476.6
10 538.5 531.9 525.2 518.4 511.6 504.9 498.6 492.9 488.0 484.2

106
 VITA

Ali Yilmaz was born in Akcakoca, Turkiye on September 1, 1967 to Kazim

and Emine Yilmaz.

He attended Akcakoca High School and graduated with an honors degree

III 1984. He continued his education in Mechanical Engineering at Istanbul

Technical University, graduating in the top 2% of his class in July, 1988.

He started his graduate studies at Lehigh University with a scholarship

granted by the Turkish Ministry of Education. He continued his graduate studies

with the aid of a research assistantship, obtaining a Master of Science Degree in

Mechanical Engineering in December, 1991.

107