Fuel 84 (2005) 927–937

www.fuelfirst.com

Investigations on fouling rate in convective bundles of coal-fired

boilers in relation to optimization of sootblower operation
Sylwester Kalisz*, Marek Pronobis
Institute of Power Engineering and Turbomachinery, Silesian University of Technology, Konarskiego 18, 44-100 Gliwice, Poland
Received 1 June 2004; received in revised form 15 December 2004; accepted 22 December 2004
Available online 25 January 2005

Abstract
The article deals with results of full-scale investigations on fouling in convective bundles of chosen types of coal-fired boilers. The boilers
tested were: conventional pulverized-coal fired two-pass boilers (two types), a single-pass subcritical (a tower shape) boiler and CFB boiler.
Mechanisms of deposit formation and basics of deposit modeling are shortly discussed. An own approach to predict build-up of loose
powdery sediments was developed on the basis of the results of full-scale measurements. The correlation for maximum fouling time was
obtained by statistical processing of measurement data and is applied in optimization of sootblower system operation. The obtained
correlation depends on the following variables: geometric properties of the bundle, fly ash size distribution and basic parameters of the flue
gases—all of them are easily accessible in boiler operation.
q 2005 Elsevier Ltd. All rights reserved.

Keywords: Fouling; Deposition modeling; Sootblowing optimization

1. Introduction transfer intensity. This is in comparison to ideal state of

the heating surface, when convective heat transfer coeffi-
Nowadays, the calculations of convective heat transfer in cient reaches its maximum value.
boiler technology are conducted by empirical correlations The principal factors that influence on J are:
obtained in standard conditions that seem to be considerably
different from full-scale reality. Widely used, basic equation – fouling on heat transfer surfaces,
for convective heat transfer coefficient k can be presented as – non-uniform flue gases flow distribution over convective
follows: bundles which are situated in ducts of significant
dimensions and complicated geometry,
k Z Jkt (1) – corrosion of tubing material.
where: The influence of fouling on overall heat transfer can be
very serious and lead to 60% depreciation of normal k value.
J thermal effectiveness number,
Existing methods [1] of incorporation of fouling into
kt convective heat transfer coefficient for ideal conditions,
thermal calculations of boilers allow, for limited range of
W/(m2 K).
cases, to carry out calculations for maximum state of fouling
Thermal effectiveness number allows to incorporate which corresponds to lowest achievable values of heat
into calculations some diverse factors affecting heat transfer coefficient. In the case of coal-fired boilers
equipped with sootblower system such condition practically
does not exist because cleaning action prevents convective
* Corresponding author. Address: Division of Energy and Furnace
bundle against serious fouling. Therefore, the bundle
Technology, Royal Institute of Technology, Brinellvägen 23, 10044
Stockholm, Sweden. Tel.: C46 8 205204; fax: C46 8 207681. operates permanently between minimum and maximum
E-mail address: [email protected] (S. Kalisz). thermal effectiveness number with respect to fouling.
0016-2361/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.fuel.2004.12.010
928 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937

In some cases, sediments on heat transfer surfaces of sulphates and only slight amount of SiO2 and Al2O3. The
achieve so significant sizes that the outage in boiler composition of the deposits changes with time and they
operation is needed in order to perform manual cleaning become stronger and harder.
of heat exchangers and flue gas ducts. All these processes Within the range of flue gas temperatures from about
lead to undesirable increase of operating costs of boilers. 800 8C down to the dew-point, the tubes are covered with
Implementation of sootblowers lower the scale of fouling loose powdery deposits containing more than 50% SiO2 and
phenomena and associated operational costs but on the other over 20% Al2O3. The composition of these sediments is not
hand, implementation of such system in existing boiler essentially different from that of fly ash. The effect of the
requires capital costs and additional costs linked to temperature of the tube walls on the structure and features of
increased fatigue of the bundle as a result of erosion and the deposits may be ignored. Deposits on adjacent super-
stresses. With increasing popularity of sootblowing systems heater and evaporator tubes, for which the wall temperature
in p-f boilers and their obligatory application in CFB boilers difference reaches 200 8C, are similar.
it is obvious that optimization of cleaning is of great interest. Recently, in the work [4], there is a new concept for
The results presented in the article belong to the series of deposits classification based on their chemical composition
the R&D investigations carried out in the Institute of Power and morphology. Typical shapes of deposits noticed on
Engineering and Turbomachinery and targeted on fouling, boiler heating surfaces are presented in Fig. 1. The
slagging and optimization of sootblowing programs [2,3]. convective bundles are, as a rule, covered with powdered
sediments of shape B, C and D.

2. Classification of the sediments

3. Experimental procedure
Commonly accepted classification of mechanisms gov-
erning sediments build-up is restricted to two general Investigations on fouling rate in convection bundles of
concepts, i.e. slagging in respect to solidified deposits and boilers were carried out for four types of bituminous coal-
fouling in respect to powdered ones. Furthermore, deposits fired boilers. Classical two-pass pulverized coal firing
are often classified into high, medium and low-temperature boilers OP-380 and OP-650, one-pass (tower shape) sub-
ones by virtue of their place of build-up inside the boiler. supercritical boiler BP-1150 and circulating fluidized bed
The chemical and physical properties of deposits depend (CFB) boiler OFz-230, all manufactured by Rafako [5],
to a large extent on the temperature of the flue gases. were investigated. The attention was paid on fouling rate
Investigations prove that the composition of high tempera- only in convection pass of boilers where powdered
ture deposits (flue gas temperatures from 1050 to 800 8C) sediments are observed. Such approach is justified
differs greatly from the composition of the fly ash. Hardened when one takes into account the characteristic properties
and sintered deposits contain more Fe2O3, increased amount of the sediments, i.e. relatively high resistance coefficient

Fig. 1. Typical shapes of sediments: (A) solidified, high-temperature sediments; (B) powdered, two-sided tangential sediments; (C) one-sided tangential; (D)
fly ash bridge.
 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937 929

The manholes nearby the convective bundles were

equipped with special covers with inspection windows
made with heat resistant glass (Fig. 2a). Additionally, the
cover plate was provided with special opening with stop
cap that enables periodical insertion of measuring rule.
The tubes most visible to inspection openings were also
supplied with sediments height indicator (Fig. 2b). All
these measures were undertaken in view of serious
problems of sampling of powdered specimens and were
meant to secure the specimen against blast while pulling
out from the duct that was reported by other investigators
[7]. All these efforts were made in order to secure fouling
rate measurement on site.
The total time of investigations in some cases reached
tens of hours, so that there was no intervention in normal
boiler operation. During measurements, basic boiler oper-
ational data were acquired together with samples of coal and
fly ashes collected in the electrostatic precipitators or
convection pass hoppers for further chemical analysis and
screening evaluation.

4. Investigation results

The locations within the boilers, where the samples were

collected are shown in Table 1. Investigations were
conducted at upper generating line of the tubes in the
bundle, independent to the flue gas flow direction, until the
build-up of sediments was completed. The results presented
are only the heights of the sediment wedge built on the
upper surface even though ash bridges were observed as a
result of fouling. In Table 1, geometrical properties of
investigated tube banks are represented as transverse/longi-
tudinal pitches s1/s2 and tube diameter D.
Fig. 2. Manhole cover plate with inspection windows and sediment height
indicator.
4.1. Investigations of one-pass BP-1150 pulverized
(an order of magnitude greater than one observed for coal boiler
solidified sediments, therefore influencing heat transfer
performance) and high removability [6]—both crucial for The investigations were conducted on the levels as
optimization of conventional sootblowers. shown in Fig. 3a corresponding to the levels of
An estimation of the required sootblowing frequency sootblowers, beginning with the highest one. In this type
is determined based on observation of fouling rate. of boiler, all heating surfaces are arranged in in-line
Table 1
Locations of deposit sample in investigated boilers

Boiler Surface Arrangement s1/s2a (mm) Db (mm)

BP-1150 ECO-economizer In-line 89/75 38
P1B-primary superheater In-line 89/75 38
M1-primary reheater In-line 178/100 51
P4-fourth superheater In-line 356/75 38
M2-secondary reheater In-line 356/71 51
OP-650 SRH-primary reheater Staggered 240/50 57
OP-380 ECO-economizer In-line 63.5/50 38
SRH-steam reheater In-line 85.7/76 51
OFz-230 SSH-primary superheater In-line 150/75 38
a
Transverse/longitudinal pitch of tube bank, respectively.
b
Outer diameter of the tube.
930 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937

4.2. Investigations of two-pass OP-650 pulverized

coal boiler

In this case, investigations were carried out on three

two-pass pulverized coal boilers with steam output of
650 t/h. In order to compare investigation results, the
same heating surface, namely first stage of steam reheater
(first SRH), was chosen in convection pass of each boiler
at the level of 36 m.
Such boilers are not equipped with sootblowers, neither
in the furnace nor in the convection pass, thus the
investigated surface was to be cleaned manually.
The maximum fouling state consisted of full fly ash
bridges and was observed after time exceeding 30 h, much
longer than in BP-1150 boiler case.
Fig. 4 depicts that for the three tested cases, the initial
period of fouling is characterized by very similar rate. The
experiment for the fouling rate till 60 h was performed
only for SRH No. 3, and assumption is made that the
fouling rate, as for initial period, would be the same for
the other two cases within the whole build-up time. The
height of the sediments reached 36 mm and is in fact the
height of the completed sediments situated at the upstream
side of the tubes. In comparison to the free distance
between the surfaces of two following clean tubes (in this
case 43 mm), the measured height proves that fouling
process is very asymmetric. In general, regardless of the
flue gas flow direction, for powdered sediments the
wedges observed for upper tubes surface reached greater
heights.
Initially, the one sided tangential sediments were
observed at upstream tube side and after certain

Fig. 3. Fouling build-up versus time in convection pass of tower shape

(one pass) BP-1150 boiler.

configuration. Firstly, investigated areas of separated level

were cleaned by means of sootblowers and the estimation
of the cleaning results was performed. Then, the
measurement of fouling rate, i.e. deposit growth within
time, was carried out while the rest of levels were blown
in normal regime in order to maintain required steam
superheat.
It was noticed that the build-up of the main mass of
sediments took place at the down stream side of the tubes
(one sided tangential) while at the end of the process a little
wedge at the upstream side was observed. It must be
stressed that the upstream side of the tubes in one-pass
boilers is the lower surface while in convective pass of two-
pass boilers the situation is quite opposite. Fig. 3b shows the
results of build-up in investigated boiler with zoning for Fig. 4. Fouling rate on the surface of steam reheater (SRH) of OP-650 p-f
separated heating surfaces. boiler.
 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937 931

4.3. Investigations of two-pass OP-380 pulverized

coal boiler

Next tests were performed at the heating surfaces of the

OP-380 p-f boiler on the both sides of convection duct at the
level of 27 m (Fig. 6).
On the right side of the duct, the experiments were
carried out at the surface of the steam reheater (SRH) with
its tubes arranged in parallel direction to the inspection
manhole and therefore providing the possibility to use all
measurement techniques for sediments height estimation,
i.e. sediments height indicators and measuring rule. During
tests, the two sided tangential sediments of 12 mm were
observed and many fly ash bridges were noticed. The gap
between neighboring tubes was 25 mm (for the in-line
mode).
Fig. 5. Asymmetric shape of sediments for OP-650’s first stage of steam On the left side of the flue gas duct, the experiments were
reheater—two sided tangential sediments with thicker layer at upstream conducted at the tubes of economizer (ECO) which at the
side of tube.
level of 27 m is placed in separate bypass duct. This system
period of time the sediments developed themselves was initially designed in order to control the superheat level
into two sided tangential and finally some fly ash in such a manner that flue gas may be bypassed around
bridges were noticed at the surfaces being investigated steam reheater to lower maximum steam temperatures. Its
(Fig. 5). present-day operation is limited only to boiler start-up.

Fig. 6. Lengthwise and crosswise cut of convective pass of OP-380 boiler.

932 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937

Table 2
Velocity and temperature of the flue gases in by-pass convection duct of the
OP-380 p-f boiler

Dampers status Surface: economizer, side: left, tube Power unit

diameter: DZ38 mm, pitches: output (MWe)
s1Z63.5 mm, s2Z50 mm
w (m/s) t (8C)
‘Closed’ 3.8 400 112
Open for 20% 5.4 590 112
Open for 50% 8.2 640 111
Open for 100% 9.3 680 112

Subsequent to velocity measurements, for similar power

unit output, the fouling rate was estimated for two extreme
dampers position resulting in maximum and minimum
achievable velocity of the flue gases in the bypass duct.
Results are presented in Fig. 8.
Fig. 8 shows that by almost doubling the flue gas velocity
(from approximately 4 to 8–9 m/s), the fouling rate is halved
(from approximately 2 to 1 mm/h). This behavior proves
high sensitivity of the loose powdery sediments build-up on
Fig. 7. Fouling rate at the surfaces of economizer (ECO) and steam reheater
(SRH) for OP-380 p-f coal-fired boiler.
flue gas velocity and is in agreement with considerations
presented elsewhere [8]. It can be expected that flue gas
The loops of economizer are built in a specific way with a velocity affects fouling in twofold way. There is a certain
range of velocity for which sediments are like to grow
very narrow gap between two neighboring tubes, i.e. the
followed by other velocity range for which the erosion
longitudinal pitch is characterized by very small value of
forces prevail, fouling stops and even already built
only 12 mm. Additionally, ECO loops are oriented perpen-
sediments are destroyed. The minimum gas speed for a
dicular to the inspection manhole thus excluding the usage
heat exchanger at which particulate fouling is avoided was
of height indicator and measuring rule. Fouling growth
investigated lately in Ref. [9]. Fouling experiments have
could be only observed by means of inspection widows.
been done with particles of different sizes and different
The results for both sides are shown in Fig. 7. Full ash
materials running under different gas speeds. It was found
bridges for ECO were observed after extremely short period
that as the flow speed in the heat exchanger increases, the
of time approximately 10–12 h both due to the narrow gap thickness and the surface area of the fouling layer deposited
between tubes and the minimum but constant flue gas flow over the heat exchanger tube are reduced. It was reported
in bypass duct. One must be concerned that during normal
boiler operation bypass dampers are nominally closed but a
small leakages occur and the flue gas flow exists to some
extent. Such small flows of flue gases are characterized by
low erosion inclination resulting in fairly high fouling rate.
Bypass duct gave the great opportunity for testing the
influence of flue gas velocity on fouling rate. As it was
mentioned, bypass dampers are closed during normal boiler
operation and the data presented in Fig. 7 reflect this
situation. But there still existed the chance to control the flue
gas flow through bypass duct. In order to obtain permissible
range of flue gas velocity, a series of measurements with
different dampers orientation in bypass duct were per-
formed. Measurements were done by means of cylinder
three-signal water-cooled probe with section of suction
pyrometer. The total length of the probe was 3.5 m while the
measuring length was approximately 3 m. Results of the
measurements are presented in Table 2. It can be clearly
seen that dampers characteristic is very abrupt proving that
even relatively small dampers opening, results in strong
Fig. 8. Fouling rate for two extreme flue gas velocities in bypass duct of OP-
response in flue gas flow and thus velocity.
380 p-f boiler.
 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937 933

that the critical flow velocity for spherical copper particles fouling rate on already cleaned tubes, at the rate of 0.5 mm
required to roll a copper particle of 10 mm is 10.5 m/s while per 5 min, were noticed in CFB boiler. This proves that CFB
for 50 mm particle is only 4.5/s. Existence of limiting flow boilers have high fouling rate. During investigations, one-
speed above which fouling is avoided is related to the sided tangential sediments were observed with only a few
critical flow velocity required to roll a particle resting on a zones of fly ash bridges. The height of sediments stayed
surface Its absolute value is determined by velocity that even on both sides of the convection duct.
corresponds to the particle size most likely to stick on the
heat exchanger tube.
4.5. Chemical composition of coal, fly ash and fly ash
particle size distribution
4.4. Investigations of circulating fluidized bed
boiler OFz-230 Samples of fly ash and coal were collected during all
trials for further analysis. Fly ash samples originated from
Fly ashes of circulating fluidized bed (CFB) boilers differ two sources—from electrical precipitator hoppers or
from conventional p-f boilers. Different chemical compo- directly from convection duct hoppers. The composition
sition and other particle-size distribution influence their of coal burned in the experimental period is shown in Table
ability to build sediments. The investigations were con- 3. It can be seen that characteristic properties of the coals
ducted simultaneously at both sides of convection pass of sampled from boilers being investigated do not show any
OFz-230 CFB boiler on the heating surface of the first stage distinct differences. Relatively, the worst coal is burned in
of steam superheater (SSH) at the level of 35.3 m. Tube the case of CFB boiler.
bank arrangement was in-line. Boiler is equipped with a The analyses of fly ash samples are shown in Table 4. For
system of sootblowers and in all cases the measurements characterization purposes the X-ray fluorescence (XRF)
were started after full cleaning cycle. During investigation, method was used, that provides bulk deposit composition as
no cleaning action was performed. elemental oxides. The fly ash of OP-380 boiler is enriched in
Measurements were performed twice. Firstly, the efforts calcium due to presence of New Integrated Desulphuriza-
were put on initial stage of fouling and sampling was tion (NID) which is semi-dry DeSOx installation in bag filter
maintained at 1/2 h frequency. Much higher fouling rate in units.
comparison to ordinary p-f boiler was observed during that The fly ash with the particles diameter lesser than 30 mm
stage. Sediments height of 2–3 mm was achieved right after mostly influence the intensity of fouling rate of powdered
1 h of experiment. sediments and therefore fly ash samples were also analyzed
Second trial was planned to estimate the total time of by means of screening procedure. Obtained results are
fouling when equilibrium is reached and there is no progress presented in Fig. 10 as remaining Rx on the sieve with mesh
in fouling height. Such equilibrium state of sediments was diameter x in the form of function RxZf(x)—in so-called
observed after approximately 16 h (see Fig. 9). Very rapid Rosin–Rammler coordinates.
Fig. 10 shows that the finest fly ash particles were
characteristic for both BP-1150 and OP-380 p-f boilers and
consequently the polydispersion numbers (defined as a slope
factor of lines) which describe the variety of available
particle diameters were similar. It is also noticeable that
wide range of particle diameters—ranging from as small
particles as 25 mm and as big as 200 mm—could be observed
for each p-f boiler.

Table 3
Analysis data of bituminous coals

LHV Ash Water Sulphur Volatiles

(kJ/kg) (%) (%) (%) (%, daf)
BP-1150, sample #1 22,289 21.0 10.1 1.04 35.4
BP-1150, sample #2 20,645 21.7 11.4 0.90 39.4
BP-1150, sample #3 20,979 20.6 11.1 0.91 39.9
OP-650, sample #1 24,357 17.2 7.1 0.68 –
OP-650, sample #2 24,427 18.4 6.5 0.69 –
OP-650, sample #3 22,098 24.2 7.8 0.82 –
OP-380, sample #1 22,131 20.1 9.7 0.98 –
OP-380, sample #2 22,658 21.1 8.7 0.91 –
OFz-230, sample #1 18,179 29.9 11.5 – –
Fig. 9. Fouling rate on steam superheater (SSH) tubes of OFz-230 CFB OFz-230, sample #2 18,234 20.2 18.3 1.2 –
boiler.
934 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937
sample#3
OFz-230

8.14

2.89
0.76
2.15
8.25

0.15
21.28

10.55
44.3

0.8
EP

–
–
–
conv. pass
sample#2
OFz-230

7.21

3.35
0.56
1.80

0.66
0.13

0.20
41.06
18.02

13.24

10.78

–
conv. pass
sample#1
OFz-230

43.76
18.95

10.52
7.07

3.53
0.73
2.14
8.83
0.75
0.12
0.14
0.20
–
sample#3
OP-380

3.83

2.12

0.61
0.31
37.76
19.44

21.42

11.13
NIDa

2.3
0.5

–
–
–
sample#2
OP-380

19.95

11.82
4.58

2.16
0.65
2.11

0.64
0.27
NIDa
38.4

18.8

–
–
–
sample#1
OP-380

5.93

2.34
1.03
2.34
6.72

0.31
44.83
24.16

10.64
NIDa

0.9

–
–
–

Fig. 10. Rosin–Rammler distribution of fly ash particles (mean values for
New Integrated Desulphurization—semi-dry desulphurization method applied in bag filter (fly ash is calcified).
sample#4

each boiler type, EP-fly ashes collected in the electrostatic precipitator).

OP-650

7.05
1.96
2.14
0.41
3.18
0.53
1.20
0.37
0.11

0.10
51.97
29.04

Higher polydispersion numbers of fly ashes from CFB

EP

boiler testify for their improved dimensional homogeneity

especially in the range of particles lesser than 120 mm. The
Chemical composition of fly ash collected from electric precipitator (EP) or convective pass hopper

sample#3
OP-650

amount offly ash collected from convection dust hopper is only

6.84
2.37
2.30
0.38
2.83
0.63
1.19
0.34
0.12

0.10
50.39
28.04

approximately 5% of overall fly ash mass and is characterized

EP

by more coarse particles without any traces of particles greater

than 150 mm. The lack of coarse particles was characteristic for
sample#2
OP-650

CFB fly ash and is responsible for its weak erosion propensity.
8.17
3.40
2.58
0.61
2.73
0.81
1.11
0.37
0.12
0.11
0.10
51.19
24.96
EP
sample#1

5. Modeling of fouling
OP-650

50.73
28.40
6.30
2.17
2.23
0.44
3.01
0.61
1.22
0.42
0.20

0.11
EP

– Modeling of fouling is mostly performed by taking into

account the following range of input data:
sample#3
BP-1150

8.47
4.31
3.26
1.03
2.90

1.12
0.62
47.99
28.36

1.0
– mineralogy, chemical composition and grain-size distri-
EP

bution of coal,

–
–
–
– boiler design and its working parameters,
sample#2

– chemical composition and particle-size distribution of fly

BP-1150

8.21
4.66
3.33
1.02
3.02
1.03
1.13
0.55
47.39
28.70

ash,
Ash analysis (wt%)

EP

– fouling rate,

–
–
–
– sediments remove ability.
sample#1
BP-1150
The fouling build-up process is controlled by two inverse

7.70
3.97
3.17
0.88
3.09
0.99
1.09
0.51
47.59
28.23
EP
phenomena:

–
–
(a) particle deposition (flux M _ dep ) caused primarily by
Table 4

Mn3O4
inertial forces, thermo/electrophoresis and vapor depo-

Al2O3
Fe2O3

Na2O
MgO

P2O5
TiO2
SiO2

CaO

BaO
K2O
SO3

SrO
sition and

a
 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937 935

(b) particle detaching (flux M _ det ) caused by tangential (in hours) is proposed in present work. Statistics has been
stresses induced by gas phase flow and erosion stresses. utilized to estimate the parameters of equation in
the following form
Thus, the flux of particles depositing on exposed area A
can be described by the following equation tmax Z f ðx1 ; x2 ; .; xn Þ (5)

1 dm M _ dep M
_ where xi are the process parameters measured during full-
Z K det (2) scale investigations described in Section 4. Correlation in
A dt A A
the form of Eq. (5) is proposed for optimization of
Development of deposition within time can be described sootblower system operation.
by various relationships [10]. Often probabilistic approach Based on authors own experience and collected data it
is utilized in order to estimate collision efficiency and was assumed that the properties and dimensions of deposits,
sticking probability [11–13]. However, the most justified and consequently the maximum fouling time, are only
approach for particulate fouling resulting in loose powdery functions of:
deposits in boilers seems to be the asymptotic function to
correlate fouling data that show falling rate asymptotic 1. relative tube spacing s1Zs1/D and s2Zs2/D,
behavior [14,15]. In classic work [16], for constant, i.e. not 2. relative tube diameter D/Db,
retarded rate of fouling, detaching was assumed as 3. relative mean velocity of the flue gas in the tube bundle
w/wb,
_ det Z bm
M (3) 4. relative ash particle-size distribution R 0.03/R0.03b,
where b is the detach frequency, sK1 (R0.03—percentage of particles bigger than 30 mm, %),
Inserting Eq. (3) into Eq. (2) and integrating within (0, 0) 5. relative mean temperature of the flue gas in the tube
and (t, m) one gets so-called Kern–Seaton equation bundle T/Tb,
6. chemical constitution of the fly ash represented as B/A
m Z Mð1 K eKbt Þ (4) (base-to-acid ratio) cm.

which describes the mass of the sediments (or characteristic With this assumption in mind and through statistical
dimension—for example height; also thermal effectiveness processing of the measurement results, the following
number J) reached after time interval t by asymptotic correlation for maximum fouling time—restricted to only
deposition. It is also assumed that the stream of depositing a few independent variables—was developed:
particles (flux M _ dep ) is constant within depositing time. In

0:744
0:319
0:249
Eq. (4), parameter M indices maximal mass of the sediments w R0:03 T
tmax Z 9:01sK1:723
2 ½h
reached after sufficient period of time has passed (theoreti- wb R0:03b Tb
cally after tZN). Nevertheless, due to lack of reliable (6)
estimation of the influence parameters it is difficult to use
Eq. (4) practically. They should be estimated for real The basic values have been chosen as: DbZ32 mm,
conditions noticed in boiler convection ducts by taking into wbZ1 m/s, R0.03bZ24.5% and TbZ373 K. The correlation
consideration complicated flow patterns, temperature dis- coefficient was 0.8937 for the following range of
tributions (thermodiffusion) and flue gases composition. As parameters s2Z0.88–1.97, w/wbZ3.8–8.5, R0.03/R0.03bZ
the result of these factors of influence, despite of many 1–2.45, T/TbZ1.68–3.05. In the course of variant
investigations, there is a lack of method for practical fouling calculations, the number of variables was reduced to
calculation. only four with strongest influence on the final result. It
Literature models, similar to the Eq. (4), allow to can be stated that the influence of flue gas velocity and
describe dependence in the form of mdepositZf(t) in order to longitudinal pitch of the tube bank on the total fouling
gain intermediate fouling state after time t. It is obvious that time tmax is of primary importance. Less influence can be
asymptotic form of such models does not allow to calculate associated with particle size distribution of the fly ash and
total fouling time directly as it is by an assumption an flue gas temperature. It must be stressed here that the
infinite time. In course of full-scale experiments conducted obtained equation is, within the range of its validity, with
so far such models do not meet the reality where maximum good conformity with investigations concerning influence
fouling time tmax is precisely specified and finite. of flue gas velocity on overall fouling rate as described in
It is particularly obvious in case of fly ash bridges, which Section 4.3.
do not enlarge themselves after forming. Similar rapid The influence of fouling on heat transfer can be
retardation in fouling rate for other shapes of sediments calculated by means of thermal effectiveness number J
resulting in reaching tmax is also observed in many cases of for various forms of deposits and their properties. However,
practical interest in pulverized coal combustion systems. it is necessary to know the shape, sizes and properties of
Regarding the considerations presented above a different deposits. In fact, these values can hardly be predicted and
principle for estimation of maximum time of fouling must be investigated. To avoid problems, a simplified
936 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937

method of characterizing the effect of fouling on heat Table 5

transfer was proposed in Ref. [17] and is based on the Mean thermal effectiveness number for different sootblowing frequency
same set of independent variables like in case of Eq. (6). Sootblowing frequency Mean thermal effectiveness number, Jm
As described in Ref. [17], the thermal effectiveness number No action (Jmin) 0.5 0.6 0.7 0.8
can be expressed by the equation Each 0.125tmax 0.89 0.93 0.95 0.97

0:056

 0.25tmax 0.81 0.87 0.91 0.94
w R0:03 0:09 T K0:138 K0:394 0.33tmax 0.77 0.84 0.89 0.93
J Z 0:46sK0:111
2 cm 0.44tmax 0.73 0.81 0.87 0.92
wb R0:03b Tb
0.5tmax 0.72 0.79 0.86 0.91
(7) 0.67tmax 0.68 0.76 0.83 0.88
tmax 0.63 0.71 0.79 0.86
in the case of staggered tube banks.
1.33 tmax 0.59 0.67 0.76 0.85
For in-line tube banks, the thermal effectiveness number
may be expressed by means of the equation:

K0:104
0:185 The results for series of sootblowing time intervals tzZ
0:106 K0:044 D w 0.125tmax, 0.25tmax, 0.33tmax, 0.44tmax, 0.5tmax, 0.67tmax,
J Z 0:25s1 s2
Db wb tmax and 1.33tmax are given in Table 5.

0:432
K0:135 (8)
R0:03 T
! cm
K0:314
R0:03b Tb
7. Conclusions
where
Fe2 O3 C CaO C MgO C Na2 O C K2 O 1. Prediction of time after which deposit layer reaches its
cm Z (9)
SiO2 C Al2 O3 C TiO2 maximum height and loose powdery fouling stops in
coal-fired boilers is proposed to be performed by the
Both equations are valid for unsintered (or only slightly
following equation
sintered) powdered deposits within the range of flue gas
temperatures over dew point to about 900 8C. Consequently
0:744


K1:723 w R0:03 0:319 T 0:249
Eqs. (6)–(8) create a set of complementary formulas which tmax Z9:01s2 ½h
wb R0:03b Tb
may be directly used in optimization of sootblowing system
operation. Formula is based on statistical processing of the data
collected in full-scale tests and depends on variables easy
to estimate, even in an engineering practice. The
following variables are subject to consideration: geo-
6. Implementation of obtained results in sootblowing
metric properties of the bundle arrangement, fly ash
optimization
properties and flue gas properties. Obtained formula can
be utilized in sootblower system optimization. In this
The mean value of thermal effectiveness number
way, economically and technically justified sootblowing
between two successive sootblowing actions (time interval
frequency may be estimated.
tz) can be obtained from:
2. To maximize the sediments build-up time and decrease
Ðtz  
sootblowing frequency, it is advised to consider tube
J
J dt
Jm ðtÞ ¼ 0
min
Jmin (10) bank arrangement of relatively low longitudinal pitch.
tz High flue gas velocities and rather coarse particle size
distribution (coarse grinding quality) are also of primary
where Jm and Jmin are the mean and minimum (observed
importance. On the other hand, all these measures lead to
for maximum fouling time) values of thermal effectiveness
increased risk of fly ash erosion over a tube bank and
number, respectively.
therefore possible overall effects must be carefully
Estimation of time dependant values of mean thermal
investigated prior implementation.
effectiveness number consists of collection of fouling
growth governing data, calculations of J on the basis of
empirical correlations (7) or (8) and subsequent evaluation
of dimensionless dependence in the form of J/JminZ Acknowledgements
f(t/tmax). The minimum value of thermal effectiveness
number Jmin is reached, when no sootblowing takes place Authors would like to acknowledge financial support of
and the maximum fouling is obtained. Right after the State Committee for Scientific Research (KBN), Poland
sootblowing action, with the assumption that all deposits are within Grant No. 8T10B01317. Experimental part of the
removed, thermal effectiveness number value JZ1. Here work took place in PKE SA—Southern Power Group as
tmax is time interval characteristic for complete rebuild of well as in Opole Power Plant. Their infrastructural and
already blown deposits (Eq. (6)). personnel support is deeply acknowledged.
 S. Kalisz, M. Pronobis / Fuel 84 (2005) 927–937 937

References [9] Abd-Elhady MS, Rindt CCM, Wijers JG, van Steehoven AA,
Bramer EA, van der Meer ThH. Minimum gas speed in heat
exchangers to avoid particulate fouling. Int J Heat Mass Transfer
[1] Pronobis M. Influence of ash deposits on the heat transfer in
2004;47:3943–55.
convection boiler surfaces. Proceedings of the Joint (Technical
[10] Epstein N. Particulate fouling of heat transfer surfaces: mechanisms and
University of Ostrava and Southern Illinois University at Carbondale)
models. In: Melo LF, Bott TR, Bernardo CA, editors. Fouling science
International Conference ‘Coal, Energy and Environment’, Ostrava,
and technology. Dordrecht: Kluwer Academic; 1988. p. 143–64.
Malenovice; 1992.
[11] Huang LY, Norman JS, Pourkashanian M, Williams A. Prediction of
[2] Pronobis M. Bewertung der effektivität von Heizflächenreinigung.
ash deposition on superheater tubes from pulverized coal combustion.
International Conference—Flame Days’99, ‘Combustion and
Fuel 1996;75:271–9.
Environment’, Praha; 1999.
[12] Tomeczek J, Palugniok H, Ochman J. Modelling of deposits
[3] Pronobis M, Kalisz S. Influence of fouling and sootblowing on formation on heating tubes in pulverized coal boilers. Fuel 2004;83:
the heat transfer in convection boiler surfaces. Acta Mech Slov 213–21.
2001;3. [13] Baxter LL, DeSollar RW. A mechanistic description of ash deposition
[4] Laursen K, Frandsen F. Classification system for ash deposits based during pulverized coal combustion: predictions compared with
on SEM analyses. In: Gupta RP, Wall T, Baxter LL, editors. Impact of observations. Fuel 1993;72:1411–8.
mineral impurities in solid fuel combustion. New York: Kluwer [14] Epstein N. Elements of particle deposition onto nonporous solid
Academic/ Plenum Publishers; 1999. surfaces parallel to suspension flows. Exp Thermal Fluid Sci 1997;14:
[5] Rafako SA. Boiler Engineering Company. www.rafako.com.pl. 323–34.
[6] Wall TF, Bhattacharya SP, Baxter LL, Richards G, Harb JN. The [15] Taborek J, Aoki T, Ritter RB, Palen JW, Knudsen JG. Fouling: the
character of ash deposits and the thermal performance of furnaces. major unresolved problem in heat transfer. Chem Eng Prog 1972;
Fuel Process Technol 1995;44:143–53. 68(2):59–67.
[7] Laursen K, Frandsen F, Larsen OH. Ash deposition trials at three [16] Kern DQ, Seaton RE. A theoretical analysis of thermal surface
power stations in Denmark. Energy Fuels 1998;12:429–42. fouling. Br Chem Eng 1959;4(5):258–62.
[8] Ots A. Mechanism of ash deposit formation, corrosion and sulphur [17] Pronobis M. Weitere Erkenntnisse zum Einfluß der Verschmutzung
capture by burning calcium and chlorine containing fuels. VGB auf den Wärmeaustausch in Konvektiven Kesselrohrbündeln. VGB
PowerTech 2001;10:114–20. Kraftwerkstechnik 1986;66.