Fuel 236 (2019) 949–959

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Heat transfer and fouling performance of finned tube heat exchangers: T

Experimentation via on line monitoring
⁎
Fei-Long Wanga, Song-Zhen Tanga, Ya-Ling Hea, , Francis A. Kulackib, Yang Yua
a
Key Laboratory of Thermo-Fluid Science and Engineering of Ministry of Education, School of Energy and Power Engineering, Xi’an Jiaotong University, Xi’an, Shaanxi
710049, China
b
Department of Mechanical Engineering, University of Minnesota, 111 Church St. S. E., Minneapolis, MN 55455, United States

A R T I C LE I N FO A B S T R A C T

Keywords: An on line experimental system was set up in a heating boiler so as to reproduce dusty working environments in
Waste heat recovery heat recovery systems and to investigate heat transfer and fouling performance of different kinds of finned tube
On line experiment heat exchangers. Measurements are reported for the heat transfer coefficients, fouling and the effect of fouling on
H-type finned tube heat transfer performance of five different H-type finned tube heat exchangers. Fouling resistance and the
Heat transfer
weakened degree of heat transfer coefficient are employed to evaluate the effect of fouling on the heat transfer
Fouling
performance. Numerical simulations are also conducted on the fouling performance of each heat exchanger
bundle and compared to the measurements. Double H-type fins can slightly reduce the heat transfer performance
but also reduce fouling. Elliptical finned tubes with the same tube pitch decrease heat transfer performance and
increase fouling, and with the same relative tube pitch, significantly increase the heat transfer performance
before fouling, and can effectively reduce fouling.

1. Introduction
Ash-related issues are among the most challenging problems in
boiler systems, and they mainly can be divided into slagging and
fouling problems, which usually leads to significant deterioration of
heat transfer performance, unscheduled shut downs and even the
failure of the systems with serious economic consequence [1–4]. The
slagging problem mainly happens in the radiant section of a boiler [2].
When the gas flows to a lower temperature region, some of the va-
porised metals nucleate and condense on the heating surfaces and form
a sticky initial slagging layer. Then the coarse fly ash, with or without a
self-surface sticky layer, will deposit on the surface of the sticky initial
slagging layer by inertial impaction. Niu et al. [4] systematically re-
viewed the slagging problems in biomass combustion systems, in-
cluding slagging formation mechanisms, damages, ash utilization, and
related countermeasures. While the fouling problem mainly occurs in
the convective passes of boilers [2]. For example, in waste heat re-
covery systems, which plays an important role in reducing the energy
loss and increasing energy efficiency in many industrial processes,
fouling can be the major problem for the heat exchangers.
Extensive research has been carried out on the heat transfer and
fouling performance of heat exchangers, and research methods can be
divided into numerical methods and experimental methods. The
numerical method has its unique advantages: (1) It can reduce costs and
save time, and (2) a large number of comparative studies can be done to
determine the optimal scheme without the workload of the experi-
mentation. We have previously presented several of numerical in-
vestigations of heat transfer and fouling performance both for circular
tube heat exchangers [5–7] and finned tube heat exchangers [8–11].
Although numerical methods have obvious advantages, their reliability
and accuracy should be first verified, and validation must be based on
comparison with experimental results. Even when an optimal design is
obtained numerically, experimental testing is necessary before the
formal industrial production of the equipment.
Experimental research on the performance of heat exchangers
mainly consists of heat transfer test, fouling processes, and the effect of
fouling on heat transfer performance. Experimental methods can be
further divided into cold state, hot state and on line experimentation.
Cold state experiments mainly focus on the fouling without heat
transfer performance. Abd-Elhady et al. [12] experimentally studied
fouling performance as well as the influence of flow direction with
respect to gravity on fouling. They found that the best orientation for
lingering particulate fouling is the downward flow, where the flow
stagnation point coincides with the top point of the heat exchanger
tubes, and the growth of the fouling layer only starts from one point.
Zhan et al. [13] experimentally investigated the particle fouling

⁎
Corresponding author.
E-mail address: yalinghe@mail.xjtu.edu.cn (Y.-L. He).

https://doi.org/10.1016/j.fuel.2018.09.081
Received 3 July 2018; Received in revised form 14 August 2018; Accepted 17 September 2018
Available online 22 September 2018
0016-2361/ © 2018 Elsevier Ltd. All rights reserved.
F.-L. Wang et al.
Nomenclature
a major semi-axial diameter, mm
Ao outside overall area, m2
b short semi-axial diameter, mm
cp specific heat capacity, J/(kg·K)
D tube diameter, mm
E Young’s modulus, GPa
Fp fin pitch, mm
G fin gap, mm
H fin length, mm
k heat transfer coefficient, W/(m2·K)
Q heat transfer capacity, W
q volume flow rate, m3/s
Rf fouling resistance, m2·K/W
s cross-sectional area, m2
S1 transverse pitch, mm
S2 longitudinal pitch, mm
T temperature, K
ΔTm log mean temperature difference, K
u fluid velocity, m/s
W fin width, mm
Y Yield stress, N/m2
characteristics of wavy fin-and-tube heat exchangers, and the effects of
fin pitch, particle concentration and air velocity. They found that the
particles mostly deposit on the leading edge of the fins and the front
part of tubes.
Lee et al. [14] investigated air-side heat transfer characteristics of
spiral-type circular fin-tube heat exchangers in hot-flow experiments.
Empirical correlations for j-factors were developed for in line and
staggered fin alignment. Kawaguchi et al. [15] experimentally com-
pared the heat transfer and pressure drop characteristics of spiral fin
and serrated fin heat exchangers. Nuntaphan et al. [16] investigated the
effect of fouling on thermal performance of a spiral fin-and-tube heat
exchanger, and developed an empirical model for evaluating the
thermal resistance of the heat exchanger. Bell et al. [17] experimental
analysed the effect of fouling on the heat transfer and pressure drop
performance of a hybrid heat exchanger. They also investigated air-side
pressure drop and heat transfer performance, as well as the effect of
fouling on the plate-fin heat exchanger [18].
Both cold state and hot state experiments have been mainly the
laboratory experiments compared to on line experiments because it is
difficult to recover actual operating environments and heat exchanger
performance. As for on line experiments, Shi et al. [19] experimentally
investigated the fouling performance of helical finned tube heat ex-
changers in a circulating fluidized bed (CFB) boiler. Li et al. [20] also
conducted experiments in a CFB boiler, and studied fouling character-
istics and the effect of fouling on the heat transfer performance of H-
type finned tube heat exchangers.
The H-type finned tube heat exchanger has been proposed as a new
kind of anti-wear and anti-fouling heat exchanger. The H-type finned
tube is comprised of two separate fins welded on both side of a tube,
and forms a narrow gap between the two fins. Its appearance resembles
the letter H, hence the name “H-type finned tube”. It has been found
that fouling on finned tubes mainly occurs in the stagnation region in
front of tubes and in the recirculation region behind the tubes [8]. The
existence of the narrow gap between the fins just eliminates the support
of deposited ash, which enable easy cleaning of tube bundle by steam
(soot blowing) and mitigates fouling. Due to anti-wear and anti-fouling
performance, the H-type finned tube heat exchanger recently has been
widely used in the waste heat recovery systems. Jin et al. [21] nu-
merically investigated heat transfer and pressure drop characteristics of
the H-type finned tube heat exchanger and correlations of Nusselt
950
Fuel 236 (2019) 949–959
Greek symbols
Γ work of adhesion, J/m2
δ fin thickness, mm
ε deposit porosity
ηo overall fin surface efficiency
λ thermal conductivity, W/(m·K)
ρ density, kg/m3
σ1 relative transverse pitch
σ2 relative longitudinal pitch
υ Poisson’s ratio
φ weaken degree of heat transfer coefficient
ϕ heat balance deviation
Subscripts
0 initial clean state
ave average
f fouling
g flue gas
in inlet
out outlet
w water
number and Euler number were obtained. Chen et al. [22] experi-
mentally investigated the heat transfer and pressure drop performance
with the effects of geometric parameters. Li et al. [23] also studied the
air side heat transfer and pressure drop of H-type finned tube heat
exchangers with in line layouts, and correlations of Nusselt number and
Euler number were developed.
In previous numerical work, we found that double H-type fins can
reduce fouling compared to that of single H-type fins [24]. Han et al.
[5] also found that elliptical tubes and staggered arrangements can
reduce the fouling on heat exchangers. However few experimental
studies have investigated anti-fouling measures with the use of double
H-type fins and elliptical tubes, especially with on line experimentation
and monitoring.
In this paper, an on line fouling experimental system is described
and used for measurement of heat transfer and fouling performance of
five H-type finned tube heat exchangers. Flue gas properties and coal
ash particle properties are first analyzed, then heat transfer, fouling and
the effect of fouling on the heat transfer performance of each heat ex-
changer is determined. Numerical simulation of the fouling perfor-
mance are also conducted and compared with the experiments. Anti-
fouling measures and structures are also discussed with implications for
the design of anti-fouling heat exchangers.
2. Experimental and methods
2.1. Experimental system
The on line experimental system is shown in Fig. 1. The system is set
up on the back side of the gravity settling chamber of a heating boiler
and the front side of the bag dust collector. Fig. 2 shows the schematic
of the experimental system and includes a heat exchanger (test unit), a
flue gas circulation loop, a water circulation loop, and a test system
with data collection and monitoring.
The flue gas with fly ash particles is drawn from the boiler by an
induced draft fan and then crosses the flue gas circulation loop to
produce fouling on the heat exchanger surfaces. The red solid line in
Fig. 2 represents the water circulation loop. The hot water in the water
circulation loop comes directly from the heating pipe of the boiler
system, and is heated by the flue gas in the heat exchanger and then
sent back to the boiler system. The blue dashed line represents the test
F.-L. Wang et al.
Fig. 1. On line fouling experimental system.
system. Inlet and outlet water temperatures and inlet and outlet flue gas
temperatures through the heat exchanger are measured. A fume ana-
lyzer, electronic differential manometer and flowmeter are used to
measure the components and content of the flue gas, flue gas velocity,
and water flowrate, respectively. The flue gas path and water path
upstream and downstream of the heat exchanger are covered by 1 cm
thick insulation to minimize heat loss. The rotational speed of the in-
duced draft fan is controlled by a frequency converter with an opera-
tional range of 0–50 Hz in intervals of 0.01 Hz. The flue gas and hot
water are taken from the boiler system and are recycled through the
system, normal operation of the boiler is maintained during our ex-
periments. Compared to prior laboratory and cold flow experimental
systems, the present system can reproduce the dusty working environ-
ments seen by industrial heat exchange equipment.
2.2. Test unit
The test unit (experimental heat exchanger) is shown in Fig. 3. The
heat transfer tubes inside the heat exchanger can be different kinds and
arrangements of expanding surface tubes and are connected by metal
hoses to form the tube circuit. Transparent Polycarbonate (PC) boards
are used on both sides of the experimental heat exchanger to realize the
visual operation and facilitate the monitoring of ash accumulation on
Fig. 2. Schematic of the experimental system.
951
Fuel 236 (2019) 949–959
tube surfaces. Compared to existing testing devices installed inside a
boiler to reproduce the real dusty environment [19,20], the present
experimental heat exchanger can change the test tubes conveniently
without affecting the normal operation of the boiler.
The types of the finned tubes tested in our investigation are shown
in Fig. 4: a single H-type finned tube (SH), double H-type finned tube
(DH), single H-type finned elliptical tube (SHE), and double H-type
finned elliptical tube (DHE). The tube bundle arrangements are shown
in Fig. 5, and geometric parameters of bundles are listed in Table 1. Five
tube bundle arrangements are designated, and four of them are based
on arrangement A (S1 = 95 mm , S2 = 76 mm ) and the four types of
finned tubes (SH, DH, SHE and DHE). Bundle number 5 is based on
arrangement B (S1 = 62.5 mm , S2 = 100 mm ) and the single H-type
finned elliptical tube (SHE). It can be seen in Fig. 5 that tube bundles
based on arrangement A consist of four rows of finned tubes in the
transverse direction and six columns in the longitudinal direction.
Bundle number 5, based on arrangement B, has a smaller transverse
pitch and a larger longitudinal pitch and therefore consists of six rows
or columns of finned tubes in both transverse and longitudinal direc-
tion. This ensures that the transverse dimension is filled with test tubes
and the flow by-pass of the bundle was prevented. However the relative
transverse pitch (σ1 = S1/ D for the circular tube bundle and σ1 = S1/2b
for the elliptical tube bundle) and relative longitudinal pitch (σ2 = S2/ D
for the circular tube bundle and σ2 = S2/2a for the elliptical tube
bundle) of bundle number 5 are kept the same with bundle 1 and
bundle 2 which are based on the circular tubes. This ensure that the
minimum cross-section area and velocity are the same with each other.
For all tube bundle arrangements, only the two middle rows of finned
tubes are considered as experimental tubes, as shown in Fig. 3(a) and
Fig. 5. The other rows are treated as non-heat transfer tubes and reduce
wall effects and create a periodic flow environment.
The detailed geometric parameters of five different finned tube
bundles are listed in Table 1. In the following paragraph, we use 1#,
2#, 3#, 4# and 5# to represent bundle 1–5 for brevity.
2.3. Experimental procedure
The test tubes in the heat exchanger are easily changed without
affecting the operation of the boiler owing to the use of gas baffles in
the flow system. Once the experiment begins, the flue gas baffles are
opened, the induced draft fan is activated, and the water loop is also
activated. After the system reaches stable operation (heat balance), flue
gas velocity, components and properties are obtained. Thermocouples
are used to measure the inlet and outlet temperatures of the flue gas
F.-L. Wang et al. Fuel 236 (2019) 949–959

Flue gas outlet

Transparent PC board
Flue gas inlet
Rubber sealing ring

Metal lock master

Dummy tubes
Metal hose

Experimental tubes

Dummy tubes

Water inlet
Water outlet

(a)

(b)
Fig. 3. Test unit: (a) exploded diagram of heat exchanger, (b) photograph of heat exchanger.

H H H H

D D 2b 2b

2a
W
W

G
W

2a
G
G

S2 S2
Fp
Fp

Fp

Fp
δ

δ
δ

(a) (b) (c) (d)

Fig. 4. Four kind of finned tubes: (a) single H-type finned tube, (b) double H-type finned tube, (c) single H-type finned elliptical tube and (d) double H-type finned
elliptical tube.

side and water side of the heat exchanger. Initial measurements near
the start of a data run are used to determine the overall heat transfer
coefficient, k0, of the clean tube bundle. Subsequent measurements are
recorded in real time, and overall heat transfer coefficients, k, are then
determined for a given flue gas velocity, which is controlled by the
frequency converter and rotational speed of the fan. With heat transfer
coefficients before and after fouling, the fouling resistance of the tube
bundle and the change of fouling resistance over time can be obtained
[9]. At the conclusion of each run, the induced draft fan and water flow
are first shut off. The gas baffles in both inlet and outlet of the flue gas
circulation are then closed in order to avoid impact on the normal
operation of the boiler system.
2.4. Method of analysis
The heat transfer capacity of on the flue gas side, Qg , and water side,
Q w , are,

952
F.-L. Wang et al. Fuel 236 (2019) 949–959

A B

S1
S1
S2
S2

Fig. 5. Tube bundle arrangements.

Table 1 and cp, w are the density and specific heat capacity of flue gas and
Geometric parameters of finned tube bundles. cooling water, respectively, qw is the volume flow rate of cooling water,
Bundle A A A A B
and Tin,g , Tout,g , Tin,w and Tout,w are the inlet and outlet temperatures of
Arrangement flue gas and cooling water. In the calculation of the overall heat transfer
Bundle 1 Bundle 2 Bundle 3 Bundle 4 Bundle 5 coefficient, the heat transfer capacity is based on the average values of
Parameters SH DH SHE DHE SHE Qg and Q w ,
(mm)
Qave = (Qg + Q w )/2 (3)
Tube size (D, 38.0 38.0 50.0 × 25.0 50.0 × 25.0 50.0 × 25.0
2a × 2b) The heat balance deviation is ϕ = |Qg−Q w|/ Qave × 100% . The criteria
Fin thickness 2.0 2.0 2.0 2.0 2.0 for an acceptable heat balance and stable operation is ϕ < 5.0%.
(δ)
Fin length (H) 70.0 147.0 70.0 147.0 70.0
The overall heat transfer coefficient of the tube bundle is calculated
Fin width (W) 74.0 74.0 60.0 60.0 60.0 as,
Fin gap (G) 10.0 10.0 10.0 10.0 10.0
Qave
Fin pitch (Fp) 16.0 16.0 16.0 16.0 16.0 k=
Transverse 95.0 95.0 95.0 95.0 62.5 Ao ΔTm (4)
pitch (S1)
Longitudinal 76.0 76.0 76.0 76.0 100
where Ao is the outside overall area of the finned tube bundle, and ΔTm
pitch (S2) is the log mean temperature difference,
Relative 2.5 2.5 3.8 3.8 2.5
ΔTmax−ΔTmin
transverse ΔTm =
pitch (σ1) ln(ΔTmax /ΔTmin ) (5)
Relative 2.0 2.0 1.52 1.52 2.0
longitu- The overall heat transfer coefficient of the clean tube bundle k 0 can
dinal pitch also be expressed as [25],
(σ2)
1
k0 = 1 Ao δ w Ao 1
+ +
hi Ai λ Ai h o ηo 6
Table 2
Uncertainties of direct and indirect measured values. where h i and h o are the water side and flue gas side convective heat
transfer coefficient, Ai is the water side overall area of tube bundle, δ w
Items Parameters Uncertainties
and λ are the thickness and thermal conductivity of tube, and ηo is the
Direct measured values Tin,g , Tout,g , Tin,w , Tout,w ± 0.5 °C overall fin surface efficiency, respectively.
ρg , ρw , cp,g , cp, w ± 1.0% When the fouling occurs on the tube surface, the heat transfer
ug ± 2.0% coefficient after fouling, k , can be expressed,
qw ± 2.0%
1 1
k= = 1
Indirect measured values Qave
k
3.7%
3.9% ( 1 Ao
hi Ai
+
δ w Ao
λ Ai
+
1
h o ηo ) +
δf
λf k0
+ Rf
(7)
Rf 5.5%
where δf and λ f are the thickness and thermal conductivity of fouling
layer, and Rf is the fouling resistance, defined as the heat resistance
Table 3 difference between fouled heat transfer surface and clean surface,
Flue gas components.
1 1
Rf = −
Components Units Value k k0 (8)
Carbon dioxide (CO2) Vol % 14.4 We also use the weakened degree of heat transfer coefficient, φ , to
Oxygen (O2) Vol % 5.4 describe the effect of fouling on the heat transfer performance,
Water vapor (H2O) Vol % 5.2
Nitrogen (N2) Vol % 75.0 k 0−k
φ =
Sulfur dioxide (SO2) mg/Nm3 809.5 k0 (9)
Nitrogen oxides (NOx) mg/Nm3 1023.1
Dust mg/Nm3 4000
3. Uncertainty analysis

Qg = ρg ug scp, g (Tin,g−Tout,g ) (1) The second power transfer method [26] is used to analyze the un-
certainties of the test data. We assume that the relation between the
Q w = ρw qw cp,w (Tout,w−Tin,w ) (2)
variable y and independent variables x1, x2 , … x n can be expressed as
where s is the cross-sectional area of the test heat exchanger, ρg , ρw , cp,g y = f (x1, x2, ⋯, x n ) , and the uncertainties of each variable x i is δx i .

953
F.-L. Wang et al. Fuel 236 (2019) 949–959

(a) Sample 1 (b) Sample 2

Fig. 6. SEM pictures of deposited ash samples.

14000 2 2
δQave 1 ⎛ δQg ⎞ ⎛ δQ w ⎞
1 — SiO2 , 75% = ⎜ ⎟ + ⎜ ⎟

Qave 2 ⎝ Qg ⎠ ⎝ Qw ⎠ (11)
12000 1 2 — Al2O3 , 12%
3 — Fe2O3 , 9% The uncertainties of the dimensions of the heat exchanger are ne-
10000 glected, and the relative uncertainties of Qg and Q w in Eq. (11) are,
4 — CaCO3 , 4%
Intensity (Counts)

2 2 2 2
8000 δQg ⎛ δρg ⎞ ⎛ δug ⎞ ⎛ δcp, g ⎞ ⎛ δ ΔTg ⎞
= ⎜ ρ ⎟ + ⎜ u ⎟ + ⎜ c ⎟ + ⎜ ΔT ⎟
Qg ⎝ g ⎠ ⎝ p, g ⎠ ⎝ g ⎠ (12)
⎝ g⎠
6000 1
2 2 2 2 2 2
4000 4 4 δQ w
= ⎛⎜ δρw ⎞⎟ + ⎜⎛ δqw ⎟⎞ + ⎛ δcp, w ⎞ + ⎛ δ ΔTw ⎞
3 1 ⎜ ⎟ ⎜ ⎟

2 13 1 1 Qw ⎝ ρw ⎠ ⎝ qw ⎠ ⎝ c p, w ⎠ ⎝ ΔTw ⎠ (13)
2 1 12
23 1
2000 1
Thus the relative uncertainty of heat transfer coefficient is,

0 2 2
10 20 30 40 50 60 70 80 90
δk
= ⎛ δQave ⎞ + ⎛ δ ΔTm ⎞
⎜ ⎟ ⎜ ⎟

2 theta (°) k ⎝ Qave ⎠ ⎝ ΔTm ⎠ (14)

Fig. 7. XRD pattern of deposited ash. Similarly, the relative uncertainty of the fouling resistance Rf is,
2 2
δRf ⎛ δk ⎞ + ⎛ δk 0 ⎞
The uncertainty of variable y is, = ⎜ ⎟

Rf ⎝k ⎠ ⎝ k0 ⎠ (15)
n 2
∂y Table 2 shows the uncertainties of the direct and indirect measured
δy = ∑ ⎛ ∂x⎜ δx i ⎞
⎟

i=1 ⎝ i ⎠ (10) results. It can be seen the relative uncertainties of the heat transfer
capacity, Qave , heat transfer coefficient k and fouling resistance Rf are
The relative uncertainty of heat transfer capacity can be expressed, within 3.7%, 3.9%, and 5.5%, respectively.

(c) After fouling (top view)

Top view

(a) Before fouling (b) After fouling (back view)

Fig. 8. Fouling results on Bundle 1.

954
F.-L. Wang et al. Fuel 236 (2019) 949–959

55 55
Experimental data Experimental data
Fitting curve
Heat transfer coefficient (W⋅m-2⋅K-1)

Heat transfer coefficient (W⋅m-2⋅K-1)

Fitting curve
50 50

k0
45 45

40 k 40

35 35

30 30
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Time (min) Time (min)
(a) bundle 1 (b) bundle 2
60 60
Experimental data Experimental data
Fitting curve
Heat transfer coefficient (W⋅m-2⋅K-1)

Heat transfer coefficient (W⋅m-2⋅K-1)

Fitting curve

50 50

40 40

30 30

20 20
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Time (min) Time (min)
(c) bundle 3 (d) bundle 4
75
Experimental data
Fitting curve
Heat transfer coefficient (W⋅m-2⋅K-1)

70

65

60

55

50
0 20 40 60 80 100 120
Time (min)
(e) bundle 5
Fig. 9. Monitoring results of heat transfer coefficient for each bundle.

955
F.-L. Wang et al. Fuel 236 (2019) 949–959

70 Table 4
k0 Properties of fly ash particles in simulation.
Heat transfer coefficient (W⋅m-2⋅K-1)

60 k Properties Symbol Value Units

Density of particles ρp 2900 kg·m−3

50 Young’s modulus E 124 GPa
Poisson’s ratio υ 0.097
Yield stress Y 3.0 × 109 N·m−2
40 Work of adhesion Γ 0.13 J·m−2
Deposit porosity ε 0.54
Thermal conductivity of deposit λf 0.3 W·m−1·K−1
30

20 30
Experiment
10 Rosin Rammler
25

Mass fraction (%)

1# 2# 3# 4# 5# 20
Finned tube bundle
Fig. 10. Comparison results of heat transfer coefficient before and after fouling 15
for each bundle.

10
4. Results and discussion

4.1. Flue gas properties 5

The properties of the flue gas are important to the heat transfer
performance of heat exchangers. The fume analyzer is applied to 0
1 3 5 7 9 11 13 15 17 19
measure the components and content of flue gas, and the results are
listed in Table 3. Since the four main components of flue gas are CO2, Particle diameter (ȝm)
O2, H2O and N2, the properties (density and dynamic viscosity) of the Fig. 12. Particle size distribution modeling.
flue gas can be calculated according to the volume fraction of each main
component under the experimental temperature and pressure.
Diffraction (XRD). The XRD pattern of the deposited ash is shown in
Fig. 7. It can be seen from Fig. 7 that, the main components of the
4.2. Coal ash particle properties deposited ash include SiO2, Al2O3, Fe2O3 and CaCO3, and the mass
fraction of each component is 75%, 12%, 9% and 4%, respectively.
Fig. 6 shows the SEM analysis results of the deposited ash samples. Then the density of the fly ash can be calculated according to the vo-
The two samples are taken from different positions on the heat transfer lume fraction of each component: ρ = 2900 kg/m3.
surface. It can be seen from the figure that, there are no large difference
between the two samples. The shape of ash particles trend to be sphe-
rical, and most of the particles are around 10 μm in diameter.
The components of deposited ash are measured by means of X-Ray

16
0.004
Reduced heat transfer coefficient (%)
Fouling resistance (m2⋅K/W)

0.003 12

0.002 8

0.001 4

0.000 0
1# 2# 3# 4# 5# 1# 2# 3# 4# 5#
Finned tube bundle Finned tube bundle
(a) (b)
Fig. 11. Comparison results of fouling resistance and reduced heat transfer coefficient.

956
F.-L. Wang et al.
Deposition thickness:
(mm)
1#
2#
3#
4#
5#
Fig. 13. Comparison of experimental and numerical results.
4.3. Heat transfer and fouling characteristics
Taking Bundle number 1, for example, the photos of single H-type
finned tubes before and after fouling are taken, and the fouling results
from the back and top views of the test heat exchanger are shown in
Fig. 8. The original shiny heat exchange tubes are covered by a layer of
fouling on the surface after the experiment, and fouling occurs mainly
on the leeward side against the flue gas. Fouling results for the other
bundles are similar to those of Bundle number 1 and are not presented
here for brevity.
Fouling on the heat transfer surface will weaken heat transfer per-
formance. With all the readings and experimental data recorded in real
time during the experiment, overall heat transfer coefficients of each
bundle are continually monitored, and the results for each bundle are
shown in Fig. 9. It can be seen from Fig. 9 that due to the measurement,
instrumental and computational errors, the measured heat transfer
coefficients always display with noise ineluctably, but they all show a
downward trend. The overall heat transfer coefficient of each bundle
reduces with the fouling progress, and then gradually reaches a stable
value. As Fig. 9a shows, the heat transfer coefficient at t = 0 represents
the value for a clean tube bundle k 0 , and the stable value, k, represents
the fouled value at the conclusion of the data run.
To compare heat transfer coefficients before and after fouling for
each bundle more clearly, the comparison results are shown in Fig. 10.
It can be seen in Fig. 10 that the order of k 0 for each bundle from high to
low is: Bundle 5, 1, 3, 2 and 4, and the order for k is: Bundle 5, 1, 2, 3
and 4. Heat transfer coefficients of Bundle 5 both before and after
fouling are much higher than for the other bundles, while the difference
in both k 0 and k for Bundles 1 to 4 is not very large. Combined with the
finned tube structure and arrangements in each test bundle (Fig. 4 and
Table 1) and the experimental comparison (Fig. 10) we find that the
double H-type fins slightly reduce heat transfer performance. Elliptical
finned tubes with the same relative tube pitch can significantly increase
957
Fuel 236 (2019) 949–959
the heat transfer performance, while with the same tube pitch elliptical
tubes will slightly reduce the heat transfer performance.
To compare the fouling performance and the effect of fouling on
heat transfer, the compared results of fouling resistance Rf and wea-
kened degree of heat transfer coefficient φ are shown in Fig. 11. The
order of Rf and φ for each bundle from low to high is: Bundle 5, 2, 1, 4,
and 3. Double H-type fins reduce fouling when compared to the single
H-type fins. Elliptical finned tubes with the same relative tube pitch can
significantly reduce fouling, while under the same tube pitch, elliptical
tubes will lead to worse fouling performance.
4.4. Comparison with simulation results
The simulation of fouling characteristics on each test bundle is
conducted here and the simulation results are compared with the ex-
perimental results. In the simulation, flue gas properties, velocity and
temperature are kept the same as those in experiment. According to the
measurement results, we use the fly ash components and mass fraction
in Section 4.2, and the properties of each component are obtained from
[27–29]. Fly ash properties used in simulation are calculated according
to the volume fraction of each component, and the results are sum-
marized in Table 4.
Since the particle size is not uniform in the experiments, the Rosin-
Rammler distribution [30] is used to model the particle size distribu-
3.5
tion. The Rosin-Rammler distribution equation is Yd = e−(dp /11) , which
means the mean diameter is 11 μm and the spread parameter is 3.5. The
comparison of Rosin-Rammler distribution and experimental particle
size distribution is shown in Fig. 12, and it fits well with each other.
Structured grid is conducted for each bundle with grid in-
dependence validated. The RNG k−ε turbulence model is used to cal-
culate the fluid flow characteristics, and the discrete phase model
(DPM) is employed to trace the particle motion during the fouling
process. Fly ash particles are injected from the inlet surface each time
F.-L. Wang et al.
step with a concentration of 4 mg/m3 and a time step size of 0.001 s.
One-way coupling is employed for the particle phase. More details
about the fouling model and numerical method can be found from our
previous work [8,9], and it will not be presented here for brevity.
Fig. 13 shows the comparisons of experimental and numerical re-
sults for each test bundle. The simulation results are in good agreement
with experimental results. The fouling on each test bundle mainly oc-
curs in the stagnation region (in front of the tubes) and recirculation
region (behind the tubes).
Comparing Bundles 1 and 2, or Bundles 3 and 4, we find that the
double H-type fins slightly reduce the fouling on heat transfer surface
either for circular tube or elliptical tube. It can be explained that the
gaps between the single H-type fins produce more turbulence than
double H-type fins. More turbulence can increase heat transfer perfor-
mance, but it will also increase the fouling on heat transfer surface.
However the gaps are tiny, the effect of gaps and the differences be-
tween single and double H-type fins are relatively small.
Comparing Bundles 1 and 3, or Bundles 2 and 4, we see that with
the same tube pitch, the use of elliptical tubes leads to more fouling
than circular tubes. It is because under the same transverse pitch, the
minimum cross section area of the elliptical tubes increases and the
minimum cross section velocity decreases, which can cause the increase
of fouling.
Comparing Bundles 1, 3 and 5, it can be seen that with the same
elliptical tubes, the fouling of same relative tube pitch (Bundle 5) is
significantly reduced than Bundles 1 and 3. It is because the same re-
lative tube pitch keeps the minimum cross-section area and velocity of
Bundles 1 and 5 the same. While the elliptical tubes have a larger high
velocity region than the circular tubes, and the stagnation region and
recirculation regions are smaller, which can both reduce the fouling.
The simulations are in good agreement with experimental results,
and the simulations can well explain the reasons for the experimental
results in Section 4.3. According to the experimental and numerical
results, we find that the double H-type fins may slightly reduce heat
transfer performance and will also reduce fouling. With the same tube
pitch, changing the circular tubes to elliptical tubes will lead to a de-
crease of heat transfer performance and increase fouling and thus
should be avoided in practice. Elliptical finned tubes with the same
relative tube pitch can significantly increase heat transfer performance
before fouling and also can effectively reduce fouling.
Therefore considering their low production cost and ease of in-
stallation, the double or even multi-H-type finned tubes can be used as
anti-fouling heat exchangers in waste heat recovery systems. While in
relatively low pressure systems, since the bearing capacity of the el-
liptical tube is not as good as the circular tube, elliptical H-type finned
tubes with the same relative tube pitch can be used to obtain better heat
transfer performance in anti-fouling heat exchangers.
5. Conclusion
Experiments have been carried out on heat transfer, fouling and the
effect of fouling on heat transfer performance of five different H-type
finned tube heat exchangers. The main conclusions are,
(1) The on line fouling experimental system can highly reproduce ac-
tual dusty working environments of heat exchangers and can be
used to investigate heat transfer and fouling performance.
(2) Compared with single H-type fins, double or even multi H-type fins
can slightly reduce the heat transfer performance but will also re-
duce the fouling. Thus they can be used as the anti-fouling heat
exchangers in the waste heat recovery systems.
(3) Elliptical finned tubes with the same tube pitch will lead to the
decrease of heat transfer performance and increase of fouling, so it
should be avoided in practical application.
(4) Elliptical finned tubes with the same relative tube pitch can sig-
nificantly increase the heat transfer performance before fouling,
958
Fuel 236 (2019) 949–959
and it also can effectively reduce the fouling, so they can be used as
high thermal performance and anti-fouling heat exchangers in the
relatively low pressure systems.
Acknowledgements
The present work is supported by the National Key R&D Program of
China (2016YFB0601100). The authors would also like to thank the
Foundation for Innovative Research Groups of the National Natural
Science Foundation of China (No. 51721004), the 111 Project
(B16038), and the China Scholarship Council (award to Fei-Long Wang
for one year’s study abroad at the University of Minnesota).
References
[1] Baxter LL, Desollar RW. A mechanistic description of ash deposition during pul-
verized coal combustion – predictions compared with observations. Fuel
1993;72(10):1411–8.
[2] Li MJ, Tang SZ, Wang FL, Zhao QX, Tao WQ. Gas-side fouling, erosion and corrosion
of heat exchangers for middle/low temperature waste heat utilization: a review on
simulation and experiment. Appl. Therm. Eng. 2017;126:737–61.
[3] He YL, Tang SZ, Wang FL, Zhao QX, Tao WQ. Gas-side fouling, erosion and corro-
sion of heat exchanger for middle and low temperature flue gas waste heat recovery.
Chin. Sci. Bull. 2016;61(17):1858–76.
[4] Niu Y, Tan H, Se Hui. Ash-related issues during biomass combustion: alkali-induced
slagging, silicate melt-induced slagging (ash fusion), agglomeration, corrosion, ash
utilization, and related countermeasures. Prog. Energy Combust. 2016;52:1–61.
[5] Han H, He YL, Tao WQ, Li YS. A parameter study of tube bundle heat exchangers for
fouling rate reduction. Int. J. Heat Mass Trans. 2014;72:210–21.
[6] Tang SZ, Wang FL, Ren QL, He YL. Fouling characteristics analysis and morphology
prediction of heat exchangers with a particulate fouling model considering de-
position and removal mechanisms. Fuel 2017;203:725–38.
[7] Tang SZ, He YL, Wang FL, Tao YB. Parametric study on fouling mechanism and heat
transfer characteristics of tube bundle heat exchangers for reducing fouling con-
sidering the deposition and removal mechanisms. Fuel 2018;211:301–11.
[8] Wang FL, He YL, Tong ZX, Tang SZ. Real-time fouling characteristics of a typical
heat exchanger used in the waste heat recovery systems. Int. J. Heat Mass Transfer
2017;104:774–86.
[9] Wang FL, He YL, Tang SZ, Tong ZX. Parameter study on the fouling characteristics
of the H-type finned tube heat exchangers. Int. J. Heat Mass Transfer
2017;112:367–78.
[10] He YL, Han H, Tao WQ, Zhang YW. Numerical study of heat-transfer enhancement
by punched winglet-type vortex generator arrays in fin-and-tube heat exchangers.
Int. J. Heat Mass Transfer 2012;55(21–22):5449–58.
[11] Han H, He YL, Li YS, Wang Y, Wu M. A numerical study on compact enhanced fin-
and-tube heat exchangers with oval and circular tube configurations. Int. J. Heat
Mass Transfer 2013;65:686–95.
[12] Abd-Elhady MS, Rindt CCM, van Steenhoven AA. Optimization of flow direction to
minimize particulate fouling of heat exchangers. Heat Transfer Eng.
2009;30(10–11):895–902.
[13] Zhan F, Tang J, Ding G, Zhuang D. Experimental investigation on particle deposi-
tion characteristics of wavy fin-and-tube heat exchangers. Appl. Therm. Eng.
2016;99:1039–47.
[14] Lee M, Kang T, Kim Y. Air-side heat transfer characteristics of spiral-type circular
fin-tube heat exchangers. Int. J. Refrig. 2010;33(2):313–20.
[15] Kawaguchi K, Okui K, Kashi T. Heat transfer and pressure drop characteristics of
finned tube banks in forced convection – (Comparison of heat transfer and pressure
drop characteristics of serrated and spiral fins). J. Enhanc Heat Transfer
2005;12(1):1–20.
[16] Nuntaphan A, Kiatsiriroat T. Thermal behavior of spiral fin-and-tube heat ex-
changer having fly ash deposit. Exp. Therm. Fluid Sci. 2007;31(8):1103–9.
[17] Bell IH, Groll EA, König H. Experimental analysis of the effects of particulate fouling
on heat exchanger heat transfer and air-side pressure drop for a hybrid dry cooler.
Heat Transfer Eng. 2011;32(3–4):264–71.
[18] Bell IH, Groll EA. Air-side particulate fouling of microchannel heat exchangers:
experimental comparison of air-side pressure drop and heat transfer with plate-fin
heat exchanger. Appl. Therm. Eng. 2011;31(5):742–9.
[19] Shi YT, Gao M, Tang GH, Sun FZ, Tao WQ. Experimental research of CFB ash de-
position on helical finned tubes. Appl. Therm. Eng. 2012;37:420–9.
[20] Li F, Sun FZ, Shi YT, Chen YL. Experimental research of CFB ash deposition on H-
style finned tubes. Adv. Inf. Sci. Service Sci. 2013;37(2):420–9.
[21] Jin Y, Tang GH, He YL, Tao WQ. Parametric study and field synergy principle
analysis of H-type finned tube bank with 10 rows. Int. J. Heat Mass Transfer
2013;60:241–51.
[22] Chen H, Wang Y, Zhao Q, Ma H, Li Y, Chen Z. Experimental investigation of heat
transfer and pressure drop characteristics of H-type finned tube banks. Energies
2014;7(11):7094–104.
[23] Li X, Zhu D, Sun J, Mo X, Yin Y. Air side heat transfer and pressure drop of H type
fin and tube bundles with in line layouts. Exp. Therm. Fluid Sci. 2018;96:146–53.
[24] Wang FL, He YL, Tong ZX, Tao WQ. Numerical study of heat transfer and fly ash
deposition characteristics for two kinds of h-type finned tubes. Therm. Fluids Eng.
F.-L. Wang et al. Fuel 236 (2019) 949–959

Summer Conf. 2016. [28] Shackelford JF, Alexander W. CRC Materials Science and Engineering Handbook.
[25] Tao WQ. Numerical Heat Transfer. second ed. Xi’an, China: Xi’an Jiaotong CRC Press; 2000.
University Press; 2001. [29] Sun C, Berg JC. Effect of moisture on the surface free energy and acid–base prop-
[26] Moffat RJ. Describing the uncertainties in experimental results. Exp. Therm. Fluid erties of mineral oxides. J. Chromatogr. A 2002;969(1):59–72.
Sci. 1988;1(1):3–17. [30] ANSYS FLUENT 18.0 Theory Guide. ANSYS Inc 2017.
[27] Lide DR. CRC Handbook of Chemistry and Physics. CRC Press; 2004.

959