Article

An Optimization Study on Soot-Blowing of Air

Preheaters in Coal-Fired Power Plant Boilers
Yuanhao Shi 1,2, *, Jie Wen 1, Fangshu Cui 3 and Jingcheng Wang 2,4

1 School of Ele ctrical and Control Engine e ring, North Unive rsity of China, Taiyuan 030051, China;
we njie [email protected]
2 De partment of Automation, Shanghai Jiao Tong Unive rsity, and Ke y Laboratory of System Control and

Information Processing, Ministry of Education of China, Shanghai 200240, China ; [email protected]

3 School of Data Scie nce and Te chnology, North Unive rsity of China, Taiyuan 030051, China;

[email protected] du.cn
4 Autonomous Systems and Inte lligent Control International Joint Re search Center, Xi’an Te chnological

Unive rsity, Xi’an 710021, China

* Corre spondence: [email protected]; Te l.: +86-0351-3921005
Re ce ived: 28 January 2019; Acce pted: 8 March 2019; Publishe d: 12 March 2019

Abstract: This paper presents a comprehensive approach for optimization of soot-blowing of air
preheaters in a coal-fired power plant boiler. In the method, modeling of the cleanliness factor is
firstly proposed to monitor the ash deposition status of the air preheaters. Then, the statistical fitting
of the ash fouling status is subsequently obtained to analyze the ash fouling dynamics and
assessment of optimized soot-blowing strategies. Soot-blowing strategies are finally developed to
optimize the steam consumption and heat transfer efficiency. Our methods can achieve the fouling
monitoring and soot-blowing optimization of air preheater (APH) by using the existing monitoring
data, not requiring additional special instruments and complex computing systems. The
methodology is validated with the actual operating data of a 300 MW coal-fired power plant boiler.
The results show the effectiveness of the proposed method. It can be used for the soot-blowing
optimization in most coal-fired power plant boiler with air preheaters.

Keywords: coal-fired power plant boiler; air preheater; ash fouling monitoring; cleanliness factor;
fouling dynamics; soot-blowing optimization

1. Introduction

During the past decades, concern about energy conservation and environmental protection is
growing all over the world. Coal-fired power plants are one of the heavy consumers of primary
energy and pollutant emissions sources. Therefore, the interest on control and optimization for coal -
fired power plant boiler has consequently become extremely relevant, aiming at increasing the
thermal efficiency and reducing the pollutant emissions [1–4].
Deposition of ash on heat transfer equipment has always been one of main operation concerns
in coal-fired power plant boiler [5]. The main problems caused by ash deposition include a reduction
in boiler thermal efficiency, higher CO2 emissions, incomplete combustion with an increase in CO
and NOx emissions, and corrosion caused by sulfides [6,7]. It has been estimated that ash deposition
will be a significant source of losses of energy efficiency and availability in thermal power plants that
may add up to 1 percent under typical operating condition s [8]. Therefore, deposition minimization
is an essential purpose from both environmental and economic viewpoint s [9,10].
The deposition of coal-fired power plant boiler is mainly classified into slagging and fouling
[11,12]. Slagging is takes place where the temperature is high and radiative heat transfer is the
dominant mechanism, i.e., furnace and roof superheaters. Fouling deposition is occurs when the

Energies 2019, 12, 958; doi:10.3390/en12050958 www.mdpi.com/journal/energies

Energies 2019, 12, 958 2 of 15

temperature is relatively low and the main heat transfer mechanism is convective heat transfer, i.e.,
low temperature superheaters, reheaters, economizers, and air preheaters. Ash deposition depends
on the coal mineral matter's nature strongly .
The traditional method for a coal-fired power plant boiler to reduce boiler ash deposition is soot -
blowing. There are many kinds of soot-blowing media, such as steam and acoustic wave. Generally,
steam is always used to blow away the deposition. The consumption of steam increases the heat
losses and reduces the turbines production, thereby decreas ing plant efficiency. Although frequent
operation of soot-blowing can increase heat transfer efficiency, it leads to the waste of steam, raised
maintenance loss, and tube corrosion. On the contrary, too little blowing results in ash deposition
accumulation and, as a result, reduces thermal efficiency . Soot blowers are continuously initiated
according to predefined sequences and operating experiences in many coal -fired power plants. The
soot-blowing strategies are not optimum without a proper boiler assessment in most cases. Therefore,
optimization for soot-blowing system with actual on-load cleaning is essential to improve boiler
operating performances [13]. Researchers and engineers focus on the optimization of the soot-
blowing frequency [14] and soot-blowing durations to improve thermal efficiency and reduce
combustion pollution [15].
Online monitoring of ash deposition status in different sections of boiler is the basis of soot-
blowing optimization. It can be accomplished by using particular power plant instrumentation and
online computations or a combination of both approaches . The particular power plant
instrumentation, such as heat flux meters [16,17] and acoustic pyrometry [18], can easily show the
status of the furnace’s heat absorption by providing continuous signals, while inducing a meaningful
increase in cost of installation, maintenance, and the sensor. The fuzzy models and artificial neural
network are well applied in the monitoring of ash fouling of boilers [7,19–21], but it is difficult for
real time monitoring because of great sum of calculations. A statistical model of ash deposition was
proposed by Zheng et al. in a pulverized-coal boiler by an exact simulation Monte Carlo approach
[22,23].
Air Preheater (APH) of a coal-fired power plant boiler is an important piece of engineering
equipment, which is designed to heat air with the primary objective of increasing boiler th ermal
efficiency. Nevertheless, there are neither special instruments like heat flux meters nor systematic
research in monitoring and quantification of ash fouling characteristic in APH. Yan et al. proposed a
flue gas differential pressure approach for APH ash fouling monitoring [24], there exist some
difficulties in application in some power plant due to flue gas density fluctuations caused by the coal
quality fluctuations. As a consequence of the lack of accurate fouling level monitoring, soot blower
are usually working in a nonoptimal state. There are still some coal-fired power plant boilers without
intelligent soot-blowing systems in China. The over soot-blowing state results in increased steam
consumption and accelerates the deterioration of tube surfaces; the less soot-blowing state results in
a loss of heat absorption and a decrease in the thermal efficiency of the boiler. On the other hand, due
to the fact that the quality of coal burned in the furnace is different from the designed coal quality in
some power plants in China, the ash content is higher. Moreover, the operation conditions of the
boiler are constantly changing with the fluctuation of load. Some intelligent soot -blowing systems
based on design parameters cannot track these changes well which may cause some not timel y or
excessive soot-blowing. Therefore, it is necessary to study the real-time soot-blowing optimization
strategies to improve soot-blowing efficiency.
In this paper, a comprehensive approach is proposed for optimization of soot-blowing of APH
in a coal-fired power plant boiler, aiming at increasing the boiler efficiency and decreasing the steam
consumption. The approach is initially focused on the monitoring of ash fouling of the APH based
on the dynamic energy balance approach. Then, the dynamic monitoring model is developed through
the statistical analysis of the fouling rates under different operation conditions. Lastly, soot-blowing
optimization strategies are proposed in order to minimize the steam consumption in the ash clean ing
process in APH, in which the soot-blowing frequencies and soot-blowing durations are decided. The
method can be used to guide soot -blowing operations to improve APH performance, while no
complex computing systems and additional instruments are required. The soot-blowing strategy is
Energies 2019, 12, 958 3 of 15

convenient to guide and assess the existing soot -blowing operations of boilers without an intelligent
soot-blowing system. It is also a good supplement for some power plant boilers with an intelligent
soot-blowing system.
This paper is organized as follows. Section 2 gives a brief description on case study coal -fired
power plant boiler, related air preheaters, and its soot-blowing system. In Section 3, the online
monitoring model to calculate the ash fouling level of the APH from available plant data is
summarized. Section 4 shows the monitoring results and statistical analysis about the ash fouling
level of the APH. Soot-blowing optimization strategies for durations and frequencies are discussed
in Section 5. The paper finishes with conclusions.

2. Case Study

The comprehensive approaches described in this paper have been validated with actual data
acquired from the unit 2 at QianXi power station in China. The boiler is a design of 300 MW unit load
of drum type with steam reheating. The nominal main steam production of the boiler is 909.6 t/h at
the rated operating conditions. It is a subcritical unit of natural circulation design and balanced draft,
W-shaped flame combustion. The main heat transfer str uctures of the boiler are a double arch single
furnace, three stages of superheaters, a single reheating system with high and low temperatures
reheaters, two economizers, and two air preheaters. The steam output temperature from first
superheater is 313 ℃ and the steam output temperature from the final superheater is 540 ℃. The
boiler's main relevant operation values are listed in Table 1.

Table 1. Boile r main ope ration values.

Parameters Units Value

BECR MW 300
Coal mass flow rate kg/s 35.4
Main steam mass flow rate t/h 909.6
Superheated steam pressure MPa 17.25
Superheated steam temperature °C 540
Reheated steam flow rate t/h 743.2
Reheated steam pressure MPa 3.18
Reheated steam temperature °C 540
Feed water temperature °C 270
Total air flow rate kg/s 295
BECR—Boile r Economic Continuous Rating.

The boiler uses coal as fuel. The design checked coal of the boiler is Qianxi anthracite coal. The
coal’s main properties are listed in Table 2.

Table 2. Coal analysis properties in case study powe r plant.

C ar /% Aar  /% Oar /% H ar /% Sar /% N ar /% M ar /% Qdw /kJ/kg

58.02 27.67 0.89 1.9 0.69 0.83 10 20992

C ar —carbon content as re ceived basis, Aar —ash content as received basis, Oar —oxyge n content as
re ce ived basis, H ar —hydroge n contend as re ceived basis, S ar —sulfur conte nt as re ceived basis,
N ar —nitroge n conte nt as re ceived basis, M ar —wate r conte nt as re ceived basis, Qdw —low heat
value as re ceived basis.

This paper focuses on the ash fouling monitoring and soot-blowing optimization of APH. The
regenerative APHs are used in most of the coal-fired power plants in China. There are two main types
of regenerative APHs: the rotating plate regenerative APH and the stationary plate regenerative APH.
The APHs in this study’s power plant use the rotating plate design which comprises a central rotating
plate element that is installed within a casing divided into triple-sectors that contain seals around the
Energies 2019, 12, 958 4 of 15

element. The seals allow the element to rotate through all the sectors, but minimize the gas leakage
while providing flue gas pathways and separated gas air through each sector.
The largest sector is connected to the boiler hot flue gas outlet. It is used to absorb the energy of
hot flue gas by means of the hot gas runs past the central element. The second sector is small and is
connected to the secondary air. When the APH rotates into the sectors, the secondary air is heated
before being carried to furnace for combustion. The smallest sector is also connected to the primary
air which is routed in the pulverizers and used to heat the pulverized coal and carry them into the
furnace for combustion. Thus, the total air heated in the rotating plate regenerative APH provides
three functions: One effect is to remove the moisture from the pulverized coal dust using heat from
the air. The second function is carrying the pulverized coal to the boiler burners. The last function is
to provide the primary air for combustion. The APH rotates quite slowly to first allow optimum heat
transfer from the hot exhaust gases to the element, then, as is rotates, from the element to the cooler
air in the other sectors. The schematic diagram of the APH is shown in Figure 1.

Low temperature Low temperature

gas outlet air inlet

Air Preheater

Gas
Gas Side
Side Air
Air side
side
Leakage air

High temperature High temperature

gas inlet air outlet

Figure 1. The sche matic diagram of the air pre heater.

The APHs in the case study boiler are equipped with 4 IK-AH soot blowers aiming at the on-
load ash cleaning by steam. The soot-blowing operations are mainly based on operators’ experiences
and following a fixed sequence.

3. Ash Fouling Monitoring and Soft Measurement

Monitoring and simulation of heat transfer of coal-fired power plant boiler can be undertaken
as a set of heat transfer equipment due to the different heat transfer forms and distributed structures.
APH is an independent heat recovery system which has its own separate soot-blowing system. There
is no any particular instrumentation to assess ash deposition effects on heat transfer efficiency in APH.
Only standard measurements of pressures and temperatures of primary air, secondary air , and flue
gas are available in both inlet and outlet of the APH.

3.1. Ash Fouling Monitoring

Ash fouling monitoring provides an opportunity to know the operation status of the APHs of
the boiler. It is also the basis for the optimization of soot-blowing and maintenance.
Figure 2 depicts the calculation structure. In this model, the energy transfer red from the hot gas
side to the cold air side can be calculated by means of an energy balance approach once the available
temperatures, pressures, and mass flow s are known. In actual working conditions, the working
efficiency of the APH depends on ash deposition status but also on working status, especially the
Energies 2019, 12, 958 5 of 15

load variations. Therefore, to take away from the unwanted impacts, the energy storage variations in
the mental elements are considered. The actual heat transfer coefficient (AC) is calculated by the
dynamic energy balance model. Furthermore, the theoretical heat transfer coefficient (TC) is obtained.
The cleanliness factor, which is the ratio of AC and TC, can be used to reflect the state of ash fouling.

Other related
parameters(inlet/outlet
temperatures)

Primary
Primary
air
air flow
flow

Secondary
Secondary
air Dynamic
Dynamic Thermal
Thermal
air flow
flow AC
AC
Balance
Balance Model
Model of
of APH
APH
Ash fouling index

Exhaust
Exhaust TC
TC
gas
gas flow
flow

Flue gas physical

properties Database

Figure 2. Sche matic diagram of the monitoring model.

For the air preheater, the heat transfer rate in gas side is determined as follow s

qg = mg Cg (Tgi − Tgo ) (1)

where qg is heat transfer rate of gas side, mg is mass flow rate of flue gas, Cg is the approximate
mean specific heat of flue gas, Tgi and Tgo are the gas temperatures in the APH gas inlet and outlet,
respectively.
For the air side the heat transfer rate is

qa = ma1Ca1 (Tao1 − Tai1 ) + ma 2Ca 2 (Tao2 − Tai2 ) (2)

where q a is heat transfer rate of air side, ma1 and ma 2 are mass flow rate of primary air and
secondary air, Ca1 and Ca 2 are the mean specific heat of primary air and secondary air, and
Tai1 , Tao1 , Tai2 and Tao2 are air temperatures in the APH primary air and secondary air inlet and outlet,
respectively.
Traditionally, the energy released in the gas side equals the energy absorbed in the air side.
qa = qg (3)

However, the energy storage of the metal of heat transfer surfaces is always varying with the
load fluctuations. Therefore, we introduce the following equation to deal with these uncertainties in
this study.
qa = qg  q j (4)

where q j is the energy storage change of the metal of heat transfer surfaces in the air preheater.

Tj
q j = C j m j (5)
t
where C j , m j , and T j are the average specific heat capacity, mass, and temperature of the metal
of air preheater, respectively.
Due to the air leakage, Equation (1) can be rewritten as

qg = mg Cg (Tgi − Tgo ) +  hlf (6)

Energies 2019, 12, 958 6 of 15

where  is the air leakage rate and hlf is the heat absorbed by the leakage air.
Thus, the actual heat transfer coefficient of air preheater can be calculated as follow s

qr q
Kr = = a (7)
AT AT
where q r is the actual heat absorption rate of the heat transfer surface, A is the area of the heat
transfer surface, and T is the log mean temperature difference, which can be described as follow s

[(Tgi − Tao ) − (Tgo − Tai )]

T = (8)
[ln(Tgi − Tao ) / ln((Tgo − Tai )]

To a certain degree, K r can reflect the fouling level of the heat transfer surface. However,
assuming that the fouling level of the heat transfer surface is stable, with the increase of unit load and
fuel quality, K r will increase while the fouling level is unchanged. Therefore, we introduced the
cleanliness factor which is the ratio of actual heat transfer coefficient K r and theoretical heat transfer
coefficient K 0 .
The TC K 0 represents the ideal heat transfer efficiency under clean status, which is usually
calculated by the sum of theoretical radiation heat transfer coefficient  f and theoretical convection
heat transfer coefficient  d .
K0 =  f +  d (9)

Generally,
agb + 1  Tgb Tgb 
 f = 5.7  10−8 ah T 3   1 − ( )4 1−  (10)
2  T T 

 wd
 d = 0.65Cs Cz ( )
0.64
Pr
0.33
(11)
d 

where agb and a h are the blackness of the heat transfer surface and the flue gas, respectively, T
and Tgb are the temperature of the heat transfer surface and the flue gas, respectively, C s and C z
are the horizontal and vertical structure parameters of the heat transfer surface, respectively,  is
the heat conductivity coefficient of the flue gas, d is the pipe diameter of the exchanger,  is the
gas flow rate,  is the dynamic viscosity coefficient of the flue gas, and Pr is Prandtl constant.
The heat transfer in APHs is convection dominated form. Thus,
K0   d (12)

Therefore, the ash fouling status of APH can be described by Cleanliness Factor (CF) as follow s
Kr
CFaph = (13)
K0

Obviously, the value of CFaph should range from 0 to 1, with 1 corresponding to the clean status
of APH. In fact, CF w ill not be 1.0 or 0. Only in the ideal conditions will CF be close to 1.0 or 0.

3.2. Soft Measurement of Air Leakage Rate

The air leakage rate is always approximately determined based on gas inlet and outlet oxygen
analysis [25]. Firstly, the air leakage coefficient is determined as follow s
 = in −  out (14)

where  in and  out are the excess air coefficient of the APHs inlet and outlet, respectively.
Generally,
Energies 2019, 12, 958 7 of 15

21 21
in = ,  out = (15)
21 − O2
i
21 − O2o

where O2i and O2o are the oxygen content in exhaust gas of the APHs inlet and outlet, respectively.
Therefore, the air leakage rate can be calculated as·

 O2i (O2i − O2o )

= = (16)
 in ·90% 0.9·(21 − O2i )(21 − O2o )

Figure 3 shows the soft measurement results of air leakage rates in case study APHs , where the
sample frequency is 5 seconds. The air leakage condition of APH-A is better than APH-B, due to the
fact that APH-A was repaired a few days prior to testing. As mentioned above, this model can be
used in the online calculation of CF in APHs.

Figure 3. Air le akage rate measurement re sults.

4. Model Validation and Analysis

In this section, the thermal model is validated and analyzed with actual operational data
acquired from the 300 MW coal-fired power plant boiler. The process data under routine operation
has been retrieved from the distributed control system (DCS) and sample frequency of measured
Energies 2019, 12, 958 8 of 15

variables is 30 seconds. The monitoring results of CFs in APHs are depicted in Figure 4, together with
the unit load and soot-blowing signals. The calculation procedure was carried out in three days.

Figure 4. Monitoring re sults of Cle anliness Factor (CF) curve s in air pre he ater-A (APH-A) and air
pre he ater-B (APH-B), with some example s of diffe rent ope rational conditions.

It is observed that the fouling level of each APH evolves with distinguishing dynamics. The CFs
of both APHs decrease much faster during high load than low load. It is observed that soot -blowing
under high load begins when the CFs reach a lower level and stops at a higher level. The fact that the
heat transfer efficiencies of APHs are higher after soot-blowing illustrates the soot-blowing effect and
dynamics. Therefore, it shows that our methods can be efficiently used to monitor the work status of
the APHs in real-time.
The fouling level dynamics of APHs can be classified into three different cases: fouling deposited
case during low load (LL) or high load (HL) and in the soot-blowing case (SB). In most cases, the CFs
dynamics are approximated by an exponential function or straight line.
The fitting models and the actual data of CFs under each condition (LL, HL, and SB) are shown
in Figure 5. Figures 5a,b indicate the fitting models and actual fouling level evolution during LL and
HL, where the fitting functions are a straight line during LL and an exponential function during HL.
Figures 5c compare the fitting model of exponential function with the actual CFs during soot -blowing
interval. The parameters of the dynamical models in different cases are shown in Table 3. The fitting
models can be used for prediction of the fouling levels in APHs and soot -blowing optimization.

Table 3. Parame ters of the functions in Figure 6.

APH Parameters LL HL SB
f n (t ) −ax + b a exp( − dx ) (ax + b)(c − exp(− dx))
a 4.277 × 10 −5 0.688 1.374 × 10 −9
b 0.6785 - 0.01374
APH-A
c - - 49.13
d - 3.729 × 10 −5 0.06832
a 3.436 × 10 −5 0.6331 1.091 × 10 −6
b 0.6022 - 0.6366 × 10 −2
APH-B
c - - 97.1
d - 3.865 × 10 −5 0.06467
Energies 2019, 12, 958 9 of 15

(a)

(b)

(c)

Figure 5. CF curve s and fitting mode ls in diffe re nt cases: (a) LL, (b) HL, and (c) SB.
Energies 2019, 12, 958 10 of 15

5. Optimization of Soot-Blowing in APHs

In order to remove the ash deposition, a large proportion of power plant boilers in China use
superheated steam to blow the ash, called ‘soot-blowing’. In the actual soot-blowing process, the
exhaust gas temperature is usually used as an indicator whether to conduct soot-blowing. However,
there are many other factors that influence the exhaust gas temperature. It is difficult for operators to
estimate the ash deposition status by gas temperature. Soot-blowing operations are always set to
predefined frequency and fix durations in a large part of power plants boilers in China. As described
in the introduction, insufficient or excessive soot -blowing of the heat transfer surfaces is not desirable.
Thus, it is quite necessary to accurately grasp the fouling states of heat transfer surfaces and
adjust the soot-blowing operations in time. The main effects for optimization of soot -blowing are
maintaining thermal efficiency of heat transfer surfaces and reducing steam con sumptions. It is
always helpful to reduce the friction losses and improve the remaining useful life of the heat transfer
surfaces.

5.1. Optimization of Soot-Blowing Frequency

In this subsection, a soot-blowing strategy for optimizing the soot-blowing frequency is

discussed, which maximizes the net benefits of soot -blowing. Soot-blowing can reduce the fouling of
APH and increase the heat absorption accordingly. We defined them as soot-blowing benefits. The
soot-blowing losses considered in this study are the steam losses.

L fl (t ) Qsb (t )
Losses and Benefits

Bs fb (t )

top tn t
Figure 6. Sche matic diagram of soot-blowing optimization.

In Figure 6, the variable of the abscissa t in Figure 6 represents the soot-blowing duration,
while the ordinate represents the losses and benefits caused by soot-blowing operations. Thus, the
net benefits of soot-blowing can be expressed as follow s
Bn = Bs − L = f b (t ) − f l (t ) (17)

where Bn , Bs and L are the net benefits, benefits, and losses of soot-blowing, respectively.
f b (t ) and f l (t ) are the corresponding functions of benefits and losses of soot-blowing, respectively.
Therefore, the main objective of optimization for soot-blowing is to find the optimal time top to
obtain the maximum net benefits Max ( Bn ) .
The losses during one soot-blowing cycle can be expressed as follow s
t1 t1
L =  f (t )dt + Qsb =  f (t )dt + ms ts (hsi − hso ) . (18)
0 0

where t1 and ts are cycle time and soot-blowing time, respectively, and f (t ) is the loss function of
heat transfer efficiency of each heat transfer surface dedicated to the ash fouling.
The soot-blowing losses considered in this study are steam consumption. Thus, soot-blowing
losses Qsb in each soot-blowing cycle can be calculated as follow s
Energies 2019, 12, 958 11 of 15

Qsb = ms ts (hsi − hso ) (19)

where ms , ts , hso and hsi are the steam mass flow rate, soot-blowing durations, steam enthalpy of
condenser inlet, and soot-blower inlet, respectively.
The calculation results for APHs in the case study boiler are shown in Table 4.

Table 4. Optimum cycle time of the air pre heaters (APHs).

S-B Losses S-B Time Optimal Time Soot-blowing Cycle

APH Qsb
, kJ/min
ts
, min
tc
, min
tb
, min tc − tb
APH-A 2.04 × 10 5 15 766 1200 −434
APH-B 2.04 × 10 5 15 731 1200 −469

As shown in Table 4, we can clearly find that the existing soot -blowing cycles in the APHs are
too long. The soot-blowing is insufficient for all APHs.

5.2. Optimization of Soot-Blowing Durations

The next step of the optimization for soot -blowing of the APHs is to determine the soot-blowing
durations.
Figure 7 shows the soot-blowing benefits in the durations. The heat absorbed by APHs without
a soot-blowing operation can be expressed as follows
t1 + ts
Qn = A f n (t )Tdt (20)
t0

The actual heat absorbed by heat transfer surfaces with soot -blowing operation is
t1 + ts
Qb = A f s (t )Tdt (21)
t0

where t 0 is the soot-blowing start time, A is area of the heat transfer surface, T is log mean
temperature difference, f n (t ) and f s (t ) are CF functions without soot-blowing and with soot-blowing
operation, respectively. Obviously, f s (t ) = K r (t ) , and the CF curve can be approximated as an
exponential functions or straight line. The detail of fitting model under each operation conditions can
be seen in Section 4.
Heat transfer efficiency

With soot-blowing operation fs (t )

ts No soot-blowing operation
f (t )
n

t1 t
Figure 7. Soot-blowing be ne fits.

Therefore, the soot-blowing benefits can be calculated as follow s

Energies 2019, 12, 958 12 of 15

B = Q = Qb − Qn =A  t1 + ts

t0
K r (t )Tdt − 
t0
t1 + ts
f n (t )Tdt . (22)
t1 + ts
= A  T 
t1
 Kr (t ) − f n (t ) dt
Accordingly, the net benefits are as follows
t1 + ts
Bn = B − Qsb = A  T 
t1
 Kr (t ) − f n (t ) dt − msts (hso − hsi ) (23)

Then the optimal soot-blowing duration top to be solved are given as follow s

Max Bn
(24)
s.t. t1  0, t s  0

The corresponding optimal cleanliness factor CFop can be obtained as

K r (top )
CFop = (25)
K0

Table 5 shows the optimization results of the soot -blowing durations of each APH. We can see
clearly that the optimum soot-blowing duration of each APH is different, of which the fouling status
of APH-B is more serious than APH-A. This conclusion is consistent with Figure 3. We can draw the
conclusion that soot-blowing for each APH is in sufficient. It is quite necessary to increase the soot-
blowing time for the APHs in the case study boiler.

Table 5. Optimum soot-blowing durations of the APHs.

S-B Losses S-B Time Optimal Soot-Blowing Durations Difference

APH Qsb
, kJ/min
ts
, min top , min ts − top , min
APH-A 2.04 × 10 5 15 19.9 −4.9
APH-B 2.04 × 10 5 15 20.4 −5.4

From what have been mentioned above, the implementation procedure and development of
soot-blowing optimization consequently is as follows

Step 1: Ash fouling monitoring. Firstly, the ash fouling status of each APH is expressed as CFs
and monitored online.
Step 2: Fitting models parameters determination. Determine the model functions and related
parameters under different operation conditions.
Step 3: Soot-blowing optimization model calculating. Thirdly, calculate the optimal soot-
blowing cycle. Consequently, the optimal cleanliness factor CFop will be calculated by solving
Max Bn in Equation (24).
Step 4: Constraints judgments. In this step, the soot-blowing operation’s required constraints
will be considered. Is the unit load is stable? Will the CF reach its optimal degree?
Step 5: Start soot-blowing. Start soot-blowing if the conditions above are satisfied.
Step 6: Determine the optimal soot-blowing durations. Calculate the optimal soot-blowing
duration top .
Step 7: End soot-blowing and enter the next cycle. If the soot-blowing time reaches the optimal
time, top , end soot-blowing and the enter next cycle. If not, continue soot-blowing.

The methodology proposed in this section can calculate the optimum soot-blowing duration and
critical soot-blowing point for each APH real time. It is effective to assess the actual soot -blowing
operations and guide the next operations.
Energies 2019, 12, 958 13 of 15

6. Conclusions

The comprehensive approaches, which were proved theoretically and practically to be feasible
and effective in monitoring ash fouling status and optimization of soot-blowing in APHs in a coal-
fired power plant boiler, are presented in this paper. The approach combines online modeling of heat
transfer efficiency to monitor the fouling level, statistical fitting to characterize the CFs’ dynamics ,
and soot-blowing optimization aiming at steam consumption conservation. Our methods can achieve
the fouling monitoring and soot-blowing optimization of APH by using the existing monitoring data.
The method can be used to guide soot-blowing operations to improve the APH’s performance, while
no complex computing systems and additional instruments are required. The soot-blowing strategy
is convenient to guide and assess the existing soot -blowing operations of the boilers without
intelligent soot-blowing system. It is also a good supplement for some power plant boilers with an
intelligent soot-blowing system.
The validation results show that the ash fouling monitoring model can provide a good
assessment of ash fouling of APHs in a relatively economical and simple way.

Author Contributions: Y.S. and J.W. conceived the theme; Y.S. and J.W. de signed and pe rformed the nume rical
simulations; and Y.S. and F.C. analyze d the re sults and wrote the original manuscript.

Funding: This re se arch was funde d by the National Natural Scie nce Foundation of China (No. 61533013), the
ke y Program of Re search and De ve lopment of ShanXi Province (No.201703D111011), the Natural Science
Foundation of ShanXi Province (No.201801D121159), the Youth Natural Scie nce Foundation of ShanXi Province
(No.201801D221208), the Natural Scie nce Foundation of North Unive rsity of China (No.2016032 and 2017025),
and the Shaanxi Provincial Ke y Proje ct (2018ZDXM-GY-168).

Conflicts of Interest: The authors de clare no conflicts of inte rest.

Abbreviations

The following abbre viations are used in this manuscript.

B be ne fits
C spe cific he at capacity (kJ/(kg·K))
d diame te r (m)
K he at transfer coefficie nt (kJ/(m 2 ·s))
L losse s
m mass flow rate (kg/s)
Pr Planck constant (J/s)
Q e ne rgy (kJ)
q he at transfer rate (kJ/s)
T te mpe rature (°C)
Gre e k symbols
 air le akage rate (%)
 incre me nt
 e xce ss air coefficient
 he at conductivity coefficient (W/(m·K))
 gas flow rate (m/s)
 dynamic viscosity coefficie nt (m 2 /s)
Subscripts/Superscripts
0 the ore tical value s
a air side
ar as re ceived basis
g gas side
i / in inle t se ction
j me ntal se ction
l loss value s
lf le akage air
Energies 2019, 12, 958 14 of 15

n ne t value s
o / out outle t se ction
op optimal value s
r actual value s
s / sb soot-blowing ope ration
Abbre viations
BECR Boile r Economic Continuous Rating
APH air pre he ater
CF cle anline ss factor
HL high load
LL low load
SB soot-blowing

References
1. Sandbe rg, J.; Fdhila, R.B.; Dahlquist, E.; Ave lin, A. Dynamic simulation of fouling in a circulating fluidize d
biomass-fired boile r. Appl. Energy 2011, 88, 1813–1824.
2. Dı́Ez, L.I.; Corté s, C.; Arauzo, I.; Valero, A. Combustion and he at transfer monitoring in large utility boilers.
Int. J. Therm. Sci. 2001, 40, 489–496.
3. Yan, Q.; Wang, Y.; Bale že ntis, T.; Sun, Y.; Stre imikie ne , D. Ene rgy-Re lated CO 2 Emission in China’s
Provincial The rmal Ele ctricity Ge neration: Driving Factors and Possibilitie s for Abatement. Energies 2018,
11, 1096.
4. Han, Y.; Xu, C.; Xu, G.; Zhang, Y.; Yang, Y. An Improve d Fle xible Solar Thermal Ene rgy Integration Process
for Enhancing the Coal-Based Ene rgy Efficie ncy and NOx Re moval Effe ctive ness in Coal-Fire d Power
Plants unde r Diffe re nt Load Conditions. Energies 2017, 10, 1485.
5. Harding, N.S.; O’Connor, D.C. Ash de position impacts in the powe r industry. Fuel Process. Technol. 2007,
88, 1082–1093.
6. Shao, Y.; Wang, J.; Pre to, F.; Zhu, J.; Xu, C. Ash De position in Biomass Combustion or Co -Firing for
Powe r/He at Ge neration. Energies 2012, 5, 5171–5189.
7. Rome o, L.M.; Gare ta, R. Ne ural ne twork for e valuating boile r be haviour. Appl. Therm. Eng. 2006, 26, 1530–
1536.
8. Vale ro, A.; Corté s, C. Ash fouling in coal-fire d utility boile rs. Monitoring and optimization of on-load
cle aning. Prog. Energy Combust. Sci. 1996, 22, 189–200.
9. Dong, M.; Han, J.; Li, S.; Pu, H. A Dynamic Mode l for the Normal Impact of Fly Ash Particle with a Planar
Surface. Energies 2013, 6, 4288–4307.
10. Dong, M.; Li, S.; Xie , J.; Han, J. Expe rime ntal Studie s on the Normal Impact of Fly Ash Particles with Planar
Surfaces. Energies 2013, 6, 3245–3262.
11. Barre tt, R.E.; Tuckfie ld, R.C.; Thomas, R.E. Slagging and Fouling in Pulverized-Coal-Fired Utility Boilers:
Volume 1, A Survey And Analysis of Utility Data: Final Report; EPRI: Palo Alto, CA, USA, 1987.
12. Williams, A. Unde rstanding slagging and fouling during PF combustion: Gordon Couch IEA Coal
Re se arch, London, 1994, pp. 109, ISBN 92-9029-240-7, £85. Fuel 1995, 74, 1541.
13. Pe ña, B.; Te rue l, E.; Díe z, L.I. Towards soot-blowing optimization in supe rhe aters. Appl. Therm. Eng. 2013,
61, 737–746.
14. Pattanayak, L.; Ayyagari, S.P.K.; Sahu, J.N. Optimization of sootblowing fre que ncy to improve boiler
pe rformance and reduce combustion pollution. Clean Technol. Environ. Policy 2015, 17, 1897–1906.
15. Mace k, K.; Ende l, P.; Cauchi, N.; Abate , A. Long-te rm predictive maintenance: A study of optimal cle aning
of biomass boile rs. Energy Build. 2017, 150, 111–117.
16. Hu, Z.; Matovic, D. He at Flux Monitoring in Biomass -Fired Boile rs: Possible Are as of Improvement. In
Proce edings of 4th IASTED Inte rnational Confe rence on Environme ntal Management and Engine e ring,
Honolulu, HI, USA, 2009; pp. 33‒38
17. Die z, L.; Corte s, C.; Campo, A. Mode lling of pulve rize d coal boile rs: Re vie w and validation of on -line
simulation te chnique s. Appl. Therm. Eng. 2005, 25, 1516–1533.
18. Zhang, S.; She n, G.; An, L.; Li, G. Ash fouling monitoring based on acoustic pyrometry in boile r furnaces.
Appl. Therm. Eng. 2015, 84, 74–81.
Energies 2019, 12, 958 15 of 15

19. Te rue l, E.; Corté s, C.; Díe z, L.I.; Arauzo, I. Monitoring and pre diction of fouling in coal-fire d utility boilers
using ne ural ne tworks. Chem. Eng. Sci. 2005, 60, 5035–5048.
20. Pe ña, B.; Te rue l, E.; Díe z, L.I. Soft-computing mode ls for soot-blowing optimization in coal-fire d utility
boile rs. Appl. Soft Comput. 2011, 11, 1657–1668.
21. Jidong, L.; Liu, D.; Liu, G.; Kai, S. Re se arch on fuzzy mode l for the soot blowing optimization in utility
boile rs. J. Huazhong Univ. Sci. Technol. 2005, 33, 35–37.
22. Zhe ng, S.; Ze ng, X.; Qi, C.; Zhou, H. Mathe matical modeling and e xperimental validatio n of ash deposition
in a pulve rize d-coal boile r. Appl. Therm. Eng. 2017, 110, 720–729.
23. Zhe ng, S.; Ze ng, X.W.; Qi, C.B.; Zhou, H.C. Mode ling of ash de position in a pulve rize d-coal boile r by direct
simulation Monte Carlo me thod. Fuel 2016, 184, 604–612.
24. Yan, W.; Che n, B.; Liang, X.; Xue , Y.E.; Wang, L.; Zhao, B. Inve stigation on Ash Monitoring Mode l for
Rotary Air He ate r in Utility Boile r. Power Eng. 2002, 22, 1708–1710.
25. Chinese Utility Boiler Performance Test Code; GB10184-88; Chine se GB Standard: Shanghai, China, 1988.

© 2019 by the authors. Lice nsee MDPI, Base l, Switze rland. This article is an ope n access
article distribute d unde r the te rms and conditions of the Creative Commons Attribution
(CC BY) lice nse (http://creativecommons.org/licenses/by/4.0/).