Journal Pre-proof

Application of optimization method based on discretized thermal energy in

condensing heat recovery system of combined heat and power plant

Wei Chen, Wenxing Shi, Xianting Li, Baolong Wang, Yang Cao

PII: S0360-5442(20)32120-4
DOI: https://doi.org/10.1016/j.energy.2020.119013
Reference: EGY 119013

To appear in: Energy

Received Date: 10 April 2020

Revised Date: 11 August 2020
Accepted Date: 2 October 2020

Please cite this article as: Chen W, Shi W, Li X, Wang B, Cao Y, Application of optimization method
based on discretized thermal energy in condensing heat recovery system of combined heat and power
plant, Energy, https://doi.org/10.1016/j.energy.2020.119013.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2020 Elsevier Ltd. All rights reserved.

Wei Chen: Conceptualization, Methodology, Investigation, Writing-original draft
Wenxing Shi : Methodology, Supervision
Xianting Li : Conceptualization, Methodology, Writing-review & editing, Supervision
Baolong Wang : Writing-review & editing, validation
Yang Cao : Writing-review & editing

of
ro
-p
re
lP
na
ur
Jo
Manuscript for Energy

Application of optimization method based on discretized

thermal energy in condensing heat recovery system of
combined heat and power plant

Wei Chen1,2, Wenxing Shi1, Xianting Li1,*, Baolong Wang1, Yang Cao2

of
1 Department of Building Science, Beijing Key Laboratory of Indoor Air Quality Evaluation

ro
and Control, Tsinghua University, Beijing 100084, China
-p
2 Institute of Building Environment and Energy, China Academy of Building Research,
re
Beijing 100084, China
lP
na

* Corresponding author: Xianting Li

ur

Department of Building Science, School of Architecture

Jo

Tsinghua University

Beijing 100084

P.R. China

Tel: +86-10-62785860

E-mail: [email protected]

-1-
Nomenclature

A area of heat exchanger, m2

Cp specific heat, kJ·kg-1·K-1
I total number of nodes in TRn
J total number of synthesized nodes
M total number of characteristic temperatures of heat demand
m mass flow rate, kg·s-1
N total number of characteristic temperatures of heat source
n number of hours

of
Q heat flow rate, kW
S set of temperatures

ro
T temperature, °C
TR
Subscripts
-p
node temperature set within a temperature range
re
c characteristic temperature of heat source or demand
grade energy grade
lP

grid temperature grid based on synthesized nodes of temperature range

hd heat demand
na

heating space heating

hs heat source
ur

input energy input

Jo

interval interval of temperature range

pinch pinch method
recovery heat recovery
Greek characters
ΔT heat transfer temperature difference, K
η heat recovery efficiency, %
Abbreviation
CHP combined heat and power
CHR conventional heat recovery system
DTE discretized thermal energy
HTTD heat transfer temperature difference
HX heat exchanger
OHR optimized heat recovery system

2
Abstract

Recovering waste heat can improve the energy efficiency of a combined heat and

power plant. The pinch method is widely used in heat recovery systems; however, the

minimum energy grade of the remaining heat demand cannot be determined using this

method. In this study, a method for discretizing thermal energy for the heat sources

and demands is proposed and illustrated with a detailed description. The proposed

method can maximize heat recovery and minimize energy grade requirements.

Different heat transfer temperature differences could be considered in the proposed

of
method. Two different optimized heat recovery systems for a combined heat and

ro
power plant are developed based on the novel and pinch methods. The amount of heat
-p
recovered using the optimized systems are the same, with values 2.08% higher than
re
that of the conventional heat recovery systems. Compared with the pinch method
lP

based optimized system, the energy grade of the remaining heat demand could be

reduced from 61.9 °C to 58.2 °C using the proposed method.

na
ur

Keywords: heat recovery; flue gas; combined heat and power; pinch method; energy

grade; discretized thermal energy

Jo

1 Introduction

A combined heat and power (CHP) generation plant with a coal-fired boiler for

hot water is energy-efficient and widely used for space heating in northern China.

Coal-fired boilers constitute approximately 85 % of the total heating and power

generation in China [1]. However, the energy-use efficiency of a coal-fired boiler is

low because of the high exhaust gas temperature. The energy-use efficiency could be

improved by efficiently recovering the waste heat contained in the flue gas.

Researchers have proposed many heat recovery technologies to improve the

energy-use efficiency of coal-fired boilers in CHP plants. Waste heat could be used to

3
reheat steam, the circulation water of the plant, supply air, or the return water of a

space-heating network. Conventional heat recovery (CHR) technologies include,

recuperators, economizers, supply air preheaters, and condensing heat recovery

systems [2,3]. The energy-use efficiency can be improved to approximately 90 %

using recuperators, economizers, and supply air preheaters. However, the use of flue

gas energy grade in a supply air preheater is not optimal. The temperature of the flue

gas is still higher than 100 °C after going through a supply air preheater [4].

Additionally, water vapor, which is the main product of a combustion reaction, is

of
released with the flue gas at a volume percentage of 12–16 % [5]. A condensing heat

recovery system was developed to recover a portion of the latent heat in the flue gas.

ro
Li et al. [6] and Wang et al. [7] proposed that the exhaust gas temperature could be
-p
reduced to less than 50 °C using condensing heat recovery. Werbe et al. [8] and Chen
re
et al. [9] illustrated that heat recovery efficiency could reach 8 % when the return
lP

water was used as the heat source. Several energy-efficient heat recovery systems

have been proposed to increase the amount of heat recovered and decrease the
na

temperature of the exhaust gas. Wei et al. [10] proposed a heat pump heat recovery
ur

system, and Chen et al. [11] and Shang et al. [12] proposed a total heat recovery
Jo

system in which the flue gas temperature can be reduced to below 30 °C. The existing

heat recovery systems for CHP plants were designed based on the experience of

experts. However, there are multiple heat demands and sources in a CHP plant, an

optimization method, rather than expert opinion, could be used to increase heat

recovery.

As a popular optimization method, the pinch method has been widely used for

optimizing heat-exchanger networks with multiple heat demands [13-16]. Hohmann

[17] proposed the pinch method in 1971, and Linnhoff [18] developed this method

further and applied it to the design of the heat-exchanger network. The pinch method

recovers the maximum amount of waste heat based on the minimum heat transfer

temperature difference (HTTD) and can be used to retrofit and design heat-exchanger

4
networks. Gadalla [19] proposed a new graphical method for a pinch analysis

application that allows the pinch point to be determined quickly. Zhang et al. [20]

integrated the energy needed of the waste heat recovery system in several factories

and reduced the annual operation cost. The pinch method can also be used to

minimize the heat-exchanger area and the total cost of the systems. Townsend et al.

[21] assumed that the heat transfer coefficients of different heat exchangers are the

same and proposed a method to determine the heat-exchanger area based on the pinch

method. Saboo et al. [22] proposed a method that considered different heat transfer

of
coefficients to determine the heat-exchanger area more accurately. The key to

determine the minimum retrofit cost is finding the optimal HTTD. Makwana et al. [23]

ro
attempted to determine the optimal minimum HTTD using a sensitivity analysis, and
-p
Trivedi et al. [24] designed an energy recovery system that considered a dual HTTD.
re
The maximum heat recovered could be obtained based on the pinch method [25,26].
lP

However, the minimum energy grade of the remaining heat demand and the public

energy cannot be obtained under certain conditions using the pinch method. The
na

energy quantity and grade are both important in an energy recovery system. If the
ur

minimum energy grade of the remaining heat demand can be obtained, the energy
Jo

grade of public energy can be reduced. Moreover, the heat-exchanger area can be

reduced along with the energy grade of public energy. Moreover, to the best of our

knowledge, few studies have considered the lowest energy grade required in a

heat-exchanger network or the application of the pinch method to flue gas recovery in

a CHP plant.

Therefore, there is a need for an optimization method that can reduce the energy

grade of public energy requirements. This method should be suitable for systems with

multiple heat demands. Herein, an optimization method based on the discretized

thermal energy (DTE) method is proposed to maximize the energy recovered and

minimize the energy grade required of the heat recovery system. The condensing heat

recovery system of a typical CHP system is selected as the study object. The heat

5
demand values of the supply air and return water are considered in the optimized

system. Both the pinch method and DTE method are used to optimize the condensing

heat recovery system, and the results can be used for the validation of DTE method.

Finally, the energy performances of the optimal and conventional systems are

compared and analyzed.

2 Description of DTE method

The quantities and grades of thermal energy are important in an energy recovery

of
system. The heat-exchanger network is made of heat exchangers. Suppose that there

ro
are three heat sources (Sa, Sb, and Sc) and demands (Da, Db, and Dc) with different

-p
average temperatures (TSa > TSb > TSc) and (TDa > TDb > TDc), respectively. In addition,

the temperature differences ∆Ta(= TSa–TDa), ∆Tb(= TSb–TDb), and ∆Tc(= TSc–TDc) are
re
assumed to be approximately the same and sufficient for heat transfer between the
lP

source and demand. If heat grading is determined, heat can be transferred between Sa
na

and Da, Sb and Db, and Sc and Dc. However, if there is no heat grading, the heat

transfer may occur between Sa and Dc first. Then, the amount of heat that can be
ur

transferred will decrease as heat transfer between Sc and Da (or Db) and between Sb
Jo

and Da may be impossible. Traditionally, there is only a qualitative concept and no

quantitative description for thermal energy grade.

For heat transfer in a heat exchanger, there must be a certain temperature

difference between the inlet and outlet of the source or demand and HTTD between

the source and demand. The HTTDs, such as ∆Ta, ∆Tb, and ∆Tc depend on economic

performance and could be different in different projects. However, the temperature

difference between the inlet and outlet of the source or demand sides should not be

too large or too small because of the reasonable convection heat transfer coefficient

and flow resistance. Therefore, the average inlet and outlet temperature of the heat

source and heat demand, and HTTD can be used to describe the heat exchange

process, i.e., the average inlet and outlet temperatures of the heat source and demand

6
can imply the energy grade of heat source and demand. The average inlet and outlet

temperature of the heat source is taken as the grade of heat source, and the average

inlet and outlet temperature of heat demand plus HTTD is taken as the grade of heat

demand. Therefore, the heat of heat sources can be transferred to demands with the

same or lower energy grade, which means that the heat-exchanger network is an

achievable network in the real project.

Therefore, the thermal energy needs to be discretized, and the grade of energy

can be described to make the cascade utilization of energy possible. The DTE method

of
is a method of discretizing thermal energy and constructing a heat-exchanger network.

The DTE method includes a total of four processing steps: 1) discretize the thermal

ro
energy values for the heat sources and demands; 2) define the energy grades of the
-p
heat sources and demands and calculate the amount of heat; 3) match the heat sources
re
and demands and create a theoretical system; 4) simplify the heat-exchanger network
lP

and create the implemented system.

na

2.1 Method of discretizing thermal energy values for heat

ur

sources and demands

Jo

In this optimization method, discretizing the energy grades of the heat sources

and demands should consider not only the inlet and outlet of heat sources and

demands but also HTTD. The following steps are introduced to complete the energy

grade division.

1) Define the characteristic temperature. The heat sources and demands have

different heat capacity. It is hard to construct a good heat exchange process if the heat

capacity of fluid on one side of the heat exchanger varies while that on the other side

is constant. The temperature point where the heat capacity changes, named as the

characteristic temperature, is set as the boundary of the DTE. The characteristic

temperature is extremely important because it is related to the temperature of the

optimized system nodes and includes the inlet temperatures of the heat sources and
7
demands and the phase transition temperature. Additionally, the characteristic

temperature must be reflected in the system. All the characteristic temperatures for the

heat sources and demands are listed in equations (1) and (2), respectively. The lines

representing the heat sources (red line) and heat demands (blue line) can be drawn, as

shown in Fig. 1.

, = , , , , , ,⋯, , , ,⋯ , , , (1)

, = , , , , , ,⋯, , , ,⋯ , , , (2)

where Sc,hs and Sc,hd are the sets of characteristic temperatures of the heat sources and

demands, respectively; Tc,hs,n represents the nth characteristic temperature of the heat

of
source (°C); Tc,hd,m represents the mth characteristic temperature of the heat demand

ro
(°C); N and M are the total numbers of characteristic temperatures of the heat sources

and heat demands, respectively.

-p
re
2) Define HTTD. The HTTD ensures that the heat from the source can be
lP

transferred to the demand side. A suitable HTTD should be defined to ensure that the

system can be implemented in a real project. In this case, the characteristic

na

temperatures of the heat demands should be revised based on the HTTD. The set of
ur

revised characteristic temperatures of the heat demands can be expressed, as in

Jo

equation (3). The HTTD may be different if two or more heat demands exist.

, = , , +∆ , , , + ∆ ,⋯, , , + ∆ ,⋯ , , +∆ (3)

3) Divide the temperature into several ranges. Based on the revised characteristic

temperature, the temperature ranges can be defined. The synthesized characteristic

temperature comprises the revised characteristic temperature of heat demands and

sources, as in equation (4). The synthesized characteristic temperatures should be

ranked from the minimum to maximum. Therefore, (N + M – 1) temperature ranges

exist if every synthesized characteristic temperature is different. The synthesized

characteristic temperatures and temperature ranges are shown by the synthesized line

(black line) in Fig. 2.

= , ∪ , = , , , ,⋯, , ,⋯ , (4)

8
where Sc represents the set of synthesized characteristic temperatures of the heat

sources and demands.

4) Create the parameters for the nodes of each temperature range. The

temperature is divided into several smaller ranges by defining a suitable interval. For

a reasonable convection heat transfer coefficient and flow resistance, the temperature

difference between the two adjacent characteristic temperatures should be much larger

than the temperature difference between the inlet and outlet of the normal heat

exchanger. The temperature interval should be defined in each temperature range

of
based on the heat exchange fluids. The temperature interval should be achievable and

reasonable for a single heat-exchanger unit. Finally, the set of node temperatures in

ro
each temperature range, which includes I elements, can be obtained using equation (5).
-p
All the node temperatures of the heat sources and demands can be obtained by
re
combining all the temperature ranges. The set of synthesized node temperatures
lP

should be ranked from the minimum to the maximum temperature, as shown in

equation (6). Temperature discretizing is the basis of DTE.

na

= , , , + , ,⋯, , + ∙ , ,⋯, , + !" − 1% , ,

ur

(5)
= ∪ ∪ ⋯∪ ∪ ⋯∪ =
Jo

& '

& , , & , ,⋯, & ,( , ⋯ & ,) , (6)

where TRn is the node temperature set within the nth temperature range; Tinterval,n is

the temperature interval in the nth temperature range (K); Tc,n+i·Tinterval,n ≤ Tc,n+1;

Tgrid,j is the temperature of the jth synthesized node (°C); I is the total number of nodes

in TRn; J is the total number of synthesized nodes.

9
 of
Fig. 1 Theoretical schematics of discretized thermal energy for heat sources and

demands

ro
-p
2.2 Energy grades and amount of heat in each energy grade
re
The primary goal of this method is to match the energy grades of the heat
lP

sources and demands to construct an optimized system. Two important elements of

na

the basic DTE are the thermal energy grade and quantity. Based on the method above,

the average temperature of two adjacent nodes is the energy grade of heat sources and
ur

demands. The quantity of heat under a certain energy grade is the sum of the heat of
Jo

all fluids between these two nodes. Subsequently, the system could be constructed

after matching energy grades.

The energy grade is defined as the average temperature of two adjacent nodes if

the specific heat capacity of the fluid is constant and is calculated using equation (7):

& , =+ & , + & , ,⁄2 , + 1 ∈ !1, 1%, (7)

where Tgrade,i represents the energy grade of the thermal energy with temperatures

from Tgrid,i to Tgrid,i+1.

The heat flow rates of the heat sources and demands of each energy grade are

defined as the heat flow rates that must be released when the fluid is cooled or heated

from one node to an adjacent node and is calculated using equations (8) and (9):

10
 2 , = !3 45 % ∙ + & , − & , , , + 1 ∈ !1, 1% (8)

2 , = !3 45 % ∙ + & , − & , , , + 1 ∈ !1, 1% (9)

where Qhs,i and Qhd,i are the heat flow rates of the heat sources and demands,

respectively, when the energy grade is Tgrade,i (kW); mhs and mhd are the mass flow

rates of the heat sources and demands, respectively (kg•s-1); and Cphs and Cphd are the

specific heat capacities of the heat sources and demands, respectively (kJ•kg-1•K-1).

The energy grades and heat flow rates of the heat sources and demands can be

presented in the same figure, as shown in Fig. 2. The X-axis shows the energy grade,

of
and Y-axis shows the heat flow rates of the heat sources and demands. The parameter

ro
at the top of the figure is the node temperature of each energy grade. According to the
-p
definitions of Tgrid and energy grade, if the energy grades of the heat sources and
re
demands are the same, the heat can be directly transferred from the source to the
lP

demand side. Furthermore, if the heat sources have higher energy grades than the heat

demands, the heat can be transferred from the heat source to the demand side. Based
na

on this feature, energy grade matching and heat transfer can be designed for different

energy grades.
ur
Jo

11
 Fig. 2 Distribution diagram of discretized thermal energy

2.3 Energy grade matching method

The energy grade matching of the heat exchanger is based on the utilization of an

energy grade cascade. Since the effect of the countercurrent heat exchange is better

than that of the downstream heat exchange [27], the former can reduce the mean

temperature difference of the heat transfer and fully utilize the energy grade. The

energy grade matching method is as follows:

of
The heat demands with the highest energy grade, Qhd,i, should be satisfied

ro
preferentially. The heat sources with the highest energy grade, Qhs,i, should also be

-p
used preferentially. The remaining heat sources of an energy grade can be used to

satisfy the heat demands of a lower energy grade. If the heat demands of an energy
re
grade are not satisfied, then public energy is needed. Based on the algorithm above,
lP

the energy grade of heat sources can be fully used because of which, the lowest
na

energy grade of remaining heat demand can be obtained. The algorithm is shown in

Fig. 3.
ur
Jo

12
 of
ro
-p
Fig. 3 Energy grade matching of heat sources and demands
re
lP

However, when the heat source accumulation is less than the heat demands
na

accumulation under a certain energy grade as presented in Fig. 4, the heat source

under this energy grade should be used preferentially. Afterward, a heat source that is
ur

higher than this energy grade can be used. In this situation, the grade of the heat
Jo

source is fully used, and the energy grade of the remaining heat demands can be

reduced significantly. Based on graphics for the energy grades and heat flow rates of

the heat sources and demands, the accumulations of the heat sources and demands can

be calculated using the following equations:

&

2 , , = 6 2 ,
(10)

&

2 , , = 6 2 ,
(11)

13
 Fig. 4 Relationship between heat source and heat demand accumulations

2.4 Simplification of heat exchange network

of
ro
The energy grades of the heat sources and demands can be matched, as

-p
mentioned in 2.3. For a theoretical system based on energy grade matching, heat

exchangers based on DTE are used. Because the theoretical system is complicated, a
re
simplification method is needed to obtain a more practical final system.
lP

A simple basic example of the simplification method is shown in Fig. 5. In Fig.

na

5(a) the two circles linked together by a line represent the heat exchanger between the

heat sources and demands. The actual heat exchanger is shown in Fig. 5(b). The
ur

shunting fluid can converge based on the energy and mass conservation principles. To
Jo

simplify the heat exchange network, parallelly connected heat exchangers are first

converted into series connections. Two such serially connected heat exchangers are

then combined into a single heat exchanger. This simplification method is feasible for

a heat-exchanger network based on the proposed energy grade matching.

(a) Expression of heat exchanger

14
 (b) Actual heat exchanger

Fig. 5 Process of simplification method

Compared with the conventional design method for a heat exchange network,

discretizing the energy grades makes it possible to fully utilize the energy and energy

grades of the waste heat, to construct an optimized system with a higher energy-use

of
efficiency.

ro
3 Application of optimization method -p
re
3.1 Object description
lP

A CHP plant simultaneously generates electricity and heat to satisfy the end-use
na

demands for both electrical power and thermal energy [28]. Fig. 6 presents the

schematics of a CHP plant. Circulation water is heated, and steam is generated in the
ur

boiler. The steam is used to generate electricity in the turbine, which is separated into
Jo

several cylinder levels based on the different steam pressures. High-pressure steam

from the boiler is used to generate electricity in the high-pressure cylinder of the

turbine. Subsequently, the steam pressure decreases to a middle-pressure level after

the high-pressure cylinder. A portion of the middle-pressure steam is extracted to heat

the return water from the district heating network. The rest enters the low-pressure

cylinder and is used to generate more electricity. Finally, the exhaust steam from the

turbine is condensed in the condenser. The exhaust flue gas from the boiler is used to

heat the circulation water successively and supply air in the economizer and air

preheater. Subsequently, the flue gas is exhausted to the atmosphere after duct

collection and desulfurization. The temperature of the flue gas after the economizer is

approximately 200 °C [29,30], and reduces to a little above 90 °C after the air
15
preheater [31,32]. This portion of heat in the flue gas must be recovered to improve

energy-use efficiency.

The typical parameters for the CHP plant are selected based on the parameters

presented in [33]. The rated capacity of the boiler is 400 MW. This plant is located in

Taiyuan City in northern China, where the ambient temperature for the design of the

heating system is –10 °C [34]. The full heating load is 135 MW when the ambient

temperature is –10 °C. The return and supply water temperatures for heating are 45

and 75 °C, respectively. The return water is heated by the steam in the steam–water

of
heat exchanger. The temperature of the exhaust gas from the boiler is 381 °C, which

subsequently decreases to 200 °C after the economizer, and to 103.7 °C after the air

ro
preheater, and finally exhausts into the atmosphere via a funnel. The supply air is
-p
heated from –10 °C to 100 °C in the air preheater. The heat demands in the CHP plant
re
include heating water, supply air, and circulation water. Furthermore, the heat sources
lP

include the exhaust gas and steam from the boiler. The primary energy savings come

from recovering maximum waste heat in the flue gas. Although this type of system
na

recovers a portion of the waste heat from the flue gas, the exhaust gas temperature is
ur

still higher than 100 °C, whereas the return water temperature of the heating network
Jo

is 45 °C. Therefore, further energy savings can be achieved.

16
 Fig. 6 Energy flow chart of CHP plant [33]

3.2 Conventional heat recovery system

The particulars of the flue gas used as the waste heat source and the supply air

and heating water that form the heat demands are presented in Table 1.

Table 1 Heat source and demands of CHP

Heat source and demands Mass flow rate Inlet temperature Outlet temperature

of
(kg/s) (°C) (°C)

ro
Flue gas 155.34 200 ~

Supply air 150.8 -p –10 100

Heating water 1076.55 45 75

re
lP

The CHR system used to improve the energy-use efficiency of a CHP plant is a
na

condensing heat recovery system. The schematics, along with the detailed parameters

under the design conditions, are shown in Fig. 7. The supply air is heated from –10 °C
ur

to 100 °C, while the flue gas temperature decreases from 200 °C to 103.7 °C.
Jo

Subsequently, the return water is used as the cooling source to recover the flue gas

waste heat via the condensing heat exchanger. The temperature of the flue gas

decreases to 55 °C, which is much lower than that of the system without condensing

heat recovery. The return water is heated from 45 °C to 46.87 °C in the condensing

heat exchanger. Subsequently, it is heated to 75 °C using steam. The condensing heat

recovery efficiency, defined as the ratio of the heat recovered to the boiler energy

input, is 6.27 %.

17
 Fig. 7 Conventional condensing heat recovery system

Although this type of heat recovery can promote the energy-use efficiency of the

of
entire system, the energy grades of the heat sources and demands do not match. The

ro
lowest energy grade of the heat demand is the temperature of the ambient air. In

-p
theory, if the energy grades between the heat sources and demands match, the

temperature of the flue gas can decrease further, and more waste heat can be
re
recovered. The optimization method based on DTE proposed can be used to optimize
lP

the heat recovery system and fully utilize the energy grades.
na

3.3 Design of optimized system

ur

The waste heat cannot satisfy the full heating load. The return water will be
Jo

heated to the supply temperature by the waste heat and public energy like steam. Two

types of optimized heat recovery (OHR) systems will be presented based on the DTE

and pinch methods. The details of the optimization process with the pinch method can

be found in [16]. The two types of optimized systems used to recover the waste heat

from flue gas are denoted as follows:

OHRDTE: an optimized system based on the DTE method;

OHRpinch: an optimized system based on the pinch method.

18
3.4 Method for evaluating optimized system

The evaluation of the OHR system is crucial to illustrate the advantages and

improvements over the conventional system. The energy-saving potential is selected

as the evaluation index for the optimization and includes the amount of heat recovered

and heat recovery efficiency under the design conditions. The heat-recovery

efficiency, which is the ratio of the heat recovered Qrecovery to the boiler’s heat input

Qinput, is as follows:
:;<=>?<;@
7 = × 100%

of
8 9 (12)
:ABCDE

ro
4 Results and discussion
-p
re
This section presents the detailed design processes of the DTE and pinch methods

applied in the CHP plant. The energy performances of the OHR and conventional
lP

systems will be compared and analyzed. Finally, system designs with multiple
na

HTTDs will be discussed.

ur

4.1 Detailed design of optimized systems

Jo

4.1.1 Optimized system based on DTE method

First, the OHRDTE system was designed based on the DTE method. The

optimization process for the DTE method is shown in this section.

Fig. 8 presents the optimization processes for OHRDTE, including, discretizing

the thermal energy values of the heat sources and demands, heat amounts for the heat

sources and demands in each energy grade, a theoretical system based on the

matching method, and the final system for implementation.

Fig. 8(a) presents the discretization of the thermal energy values of the heat

sources and demands. The HTTD of the supply air and return water is 10 °C.

According to the discretization method for thermal energy, the temperature can be

19
discretized into four ranges. Based on the different fluids used for the heat sources

and demands in each range, the four temperature intervals are 10, 5, 10, and 10 K.

Fig. 8(b) presents the calculation results for the heat sources and demands for

each grade. The amount of heat in each grade is related to its heat capacity and

temperature interval. The theoretical system of the novel heat recovery system for

OHRDTE was implemented based on the results of Fig. 8(b) and the energy grade

matching method.

The theoretical system presented using the graph of the heat-exchanger network

of
is shown in Fig. 8(c). When the energy grade is 82.5 °C, the heat source accumulation

is, at first, less than the heat demand. As shown, the flow in some temperature ranges

ro
requires shunting, the theoretical system is complicated, and many heat exchangers

are required.
-p
re
The result based on the simplification method is presented in Fig. 8(d) which
lP

requires six heat exchangers, including a steam–water heat exchanger. Furthermore,

the node temperatures of the system and mass flow rate of each heat exchanger are
na

determined after simplification.

ur
Jo

(a) Division of heat grade range

20
 of
ro
(b) Heat source and demand in each energy grade

-p
re
lP
na
ur
Jo

(c) Theoretical optimized heat recovery system

(d) Final system for implementation

21
 Fig. 8 Optimization processes for OHRDTE system

Based on the DTE method, the final system for implementation (OHRDTE system)

is presented in Fig. 8 (d). The flue gas heats the supply air first and then heats a

portion of the return water from 71.4 °C to 75 °C, while the steam heats this portion

of the return water from 45 °C to 71.4 °C using the steam–water heat exchanger. As,

the flue gas temperature decreases to 85 °C, the flue gas is divided into two portions,

one heats a portion of the return water from 45 °C to 75 °C, while another heats the

of
supply air from 45 °C to 75 °C. The latent heat of the flue gas heats the supply air

from –10 °C to 45 °C. In the OHRDTE system, the amount of heat recovered is 33.44

ro
MW, whereas the public energy consumption is 118.23 MW.
-p
4.1.2 Optimized system based on pinch method
re
The OHRpinch system was designed based on the pinch method. First, the pinch
lP

location was determined. The minimum HTTD was set at 10 °C. Based on the inlet
na

and outlet temperatures of the heat source and demands, the temperature could be

divided into four intervals, as presented in Fig. 9. Streams 1, 2, and 3 (numbers inside
ur

the square) are the supply air, return water, and flue gas, respectively. The amounts
Jo

for the heat sources and demands in each temperature interval can be calculated.

Fig. 9 Temperature interval of heat source and demands in pinch method

Table 2 lists the heat flow rate differences for the heat source and heat demands

and the accumulated net heat in each temperature interval. The results show that the

22
accumulated net heat in temperature interval 3 is negative, which means at least

118.23 MW of public energy is needed. The amount of heat recovered is 33.44 MW,

which is the same as that of the DTE method. Pinch method is a design method

recognized by many researchers in the heat recovery field. Systems can be designed to

recover the maximum heat in a heat exchanger network by pinch method. Therefore,

the results verify the correctness and rationality of the DTE method to a certain extent.

The results also indicate that the pinch temperature is 45 °C on the heat-demand side

and 55 °C on the heat-source side.

of
Table 2 Heat flow rate difference and accumulated net heat

NO. ∆T CPh-CPc ∆Q Qac

ro
(K) (kW/K) (kW) (kW)

1 90 +173.1402
-p +15582.62 +15582.62
re
2 25 +21.5862 +539.66 +16122.28
lP

3 30 –4478.3928 –134351.78 –118229.5

4 55 +21.5862 +1187.24 –115310.87

na
ur

The heat-exchanger network was divided into two subnetworks bounded by the
Jo

pinch temperature. The design of these two subnetworks was carried out separately.

One of the final systems of the maximum energy recovery network based on the

energy matching criteria of the pinch method is presented in Fig. 10. The system

requires four heat exchangers, including the steam–water heat exchanger. The flue gas

heats the supply air first and then heats the water from 45 °C to 48.73 °C, while the

steam heats the return water from 48.73 °C to 75 °C in the steam–water heat

exchanger. As the temperature of the flue gas decreases to 55 °C, the latent heat of the

flue gas heats the supply air from –10 °C to 45 °C. The amount of heat recovered is

33.44 MW.

23
 Fig. 10 Final OHRpinch system for implementation

4.2 Performance analysis of optimized system

of
The amount of heat recovered for the two OHR systems and the CHR system are

ro
presented in Table 3. The amount of heat recovered from the two OHR systems are

-p
the same, i.e., 33.44 MW. Furthermore, the OHR has a heat recovery efficiency of

8.36 %, which is 2.08 % higher than that of the CHR system. The energy grades of
re
the remaining heat demands of the OHRDTE, OHRpinch, and CHR systems are 58.2 °C,
lP

61.9 °C, and 60.9 °C, respectively. The energy grade of the remaining heat demand of
na

OHRDTE is lower than those of OHRpinch and CHR. What is more, the economic

performance is very important when the exhausted temperature is reduced to lower

ur

than dewpoint temperature because of the high capital cost of the heat exchanger with
Jo

anti-corrosion. Shang et al.[12] has analyzed the economic performance of the heat

recovery system when the exhausted temperature is reduced to lower than 30 °C. The

results show that the payback period of the heat recovery system is three years relative

to the original scheme. In the optimized system of DTE method, the final exhausted

temperature of flue gas is 42.15 °C, which is much higher than that case in Shang et

al.[12]. Therefore, the economic performance of the optimized system is better than

the case in Shang et al.[12], and will not be discussed in this paper.

Table 3 Energy-saving potentials of OHR systems under design conditions

Texhaust Qrecovery ηrecovery Qpublic Energy
(°C) (MW) (%) (MW) grade of
remaining
demand
24
OHR 1 42.15 33.44 8.36 118.23 58.2 °C
OHR 2 42.15 33.44 8.36 118.23 61.9 °C
CHR 55 25.11 6.28 126.56 60.9 °C

Compared with the OHRpinch system, the reduction of the energy grade of the

remaining heat demand in the OHRDTE system brings two advantages in the energy

grade of the steam and the steam–water heat-exchanger area. 1) Although the steam

temperatures are the same in the OHRDTE and OHRpinch systems, the steam–water heat

exchanger of the former is smaller because of the larger HTTD. 2) Although the

of
steam–water heat-exchanger areas are the same in both systems, the steam

ro
temperature of the OHRDTE system will be lower.

-p
With regard to the first advantage, the steam temperatures of the three systems

are the same, i.e., 105 °C. The heat transfer coefficients of the gas–water heat
re
exchanger, gas–air heat exchanger, and steam–water heat exchanger are 100
lP

W/(m2·K), 40 W/(m2·K), and 3000 W/(m2·K), respectively. The results of a

na

comparison of the OHRDTEa and OHRpinch systems are presented in Table 4. The

HTTDs of the steam–water heat exchangers are different (45.53 °C in the OHRDTEa
ur

system and 41.77 °C in the OHRpinch system). Therefore, the steam–water

Jo

heat-exchanger area of the OHRDTEa system is 78 m2 less than that of the OHRpinch

system.

With regard to the second advantage, the areas of the steam–water heat

exchangers of the OHRpinch and OHRDTEb systems remain the same. The results of a

comparison of the OHRDTEb and OHRpinch systems are presented in Table 4. The

steam temperature can be reduced from 105 °C to 101.35 °C.

Table 4 Comparison of heat exchanger-areas and steam temperatures

HXsteam–water HXgas–water HXgas–air Tsteam

(m2) (m2) (m2) (°C)

OHRDTEa 865 4296.04 19543.57 105

OHRDTEa
OHRDTEb 943 4296.04 19543.57 101.35

25
 OHR 2 943 4201.54 9259.69 105

CHR 991.63 3132.54 3906.36 105

Based on the above analysis, more heat can be recovered by the optimized

systems than the conventional one at the expense of an increased heat transfer area.

The energy grade of the remaining heat demand can be reduced in the OHRDTE

system, reducing the steam–water heat-exchanger area or steam temperature.

of
4.3 System design with multiple HTTDs

ro
-p
The minimum HTTDs of the OHRDTE and OHRpinch systems are the same in the

previously presented optimized systems. However, the HTTDs of the gas–water heat
re
exchanger and gas–air heat exchanger should be different because of the different heat
lP

transfer coefficients. Using the same HTTD results in a large area for the gas–air heat
na

exchanger, which may be an important factor affecting the feasibility of the optimized

system. If different HTTDs, based on the thermodynamic properties of the fluids are
ur

used, the quantity and quality of the energy-saving potential should be different.
Jo

Hence, though it is difficult to design a system based on the pinch method, it is

feasible using the DTE method.

The HTTD of the gas–air heat exchanger was set as 40 °C under the dry condition

and 25 °C under the wet condition instead of 10 °C, which caused a significant

reduction in the gas–air heat-exchanger area. Meanwhile, the HTTD of the gas–water

heat-exchanger remains at 10 °C. The heat source and demands for each energy grade

are presented in Fig. 11. The histogram of the heat demand of the supply air with

temperatures higher than 40 °C shifts to the right because of the larger HTTD and

50 °C dewpoint temperature of the flue gas.

26
 of
ro
-p
re
Fig. 11 Heat source and demand for each energy grade with different HTTDs
lP

Based on the DTE and optimization methods, OHR systems with multiple
na

HTTDs (OHRDTE,MHTTD) were designed, as shown in Fig. 12. A portion of the waste
ur

heat with a high energy grade is used to satisfy the heat demand of the supply air. The
Jo

energy grade of the remaining heat demand of the OHRDTE,MHTTD system is 58.7 °C,

which is 0.5 °C higher than that of the OHRDTE system. In other words, the energy

grade of the remaining heat demand will increase when the HTTD of any heat

exchanger is set differently.

Fig. 12 Schematics of OHR3 system with different HTTDs

27
 Table 5 presents the heat-exchanger areas in the OHRDTE,MHTTD and OHRDTE

systems. When compared to the OHRDTE systems, steam–water heat exchanger in the

OHRDTE,MHTTD has a slightly larger area when the steam temperature is 105 °C. The

gas–water heat exchanger has a much larger area, whereas the gas–air heat-exchanger

area reduced by more than 50 %. Hence, the heat-exchanger investment can be

decreased with a slight increase in the energy grade of the remaining heat demand.

Therefore, it is possible to balance the relationship between energy savings and

investment in the optimized system using multiple HTTDs.

of
Table 5 Areas of heat exchangers in OHRDTE,MHTTD system and OHRDTE system

ro
HXsteam-water HXgas-water HXgas-air Tsteam

(m2)
-p
(m2) (m2) (°C)
re
OHRDTE 865 4296.04 19543.57 105
lP

OHRDTE,MHTTD 876.51 7675.46 9259.69 105

na

5 Conclusions
ur
Jo

This paper proposed an optimization method based on the discretized thermal

energy to optimize a heat recovery system. Two heat recovery systems were

optimized based on the DTE and pinch methods, and energy-saving potential was

discussed. The following lists some of the conclusions.

1 The method of discretizing the thermal energy for the heat sources and

demands has been proposed and illustrated with detailed descriptions. The distribution

diagram of DTE shows a clear relationship between the heat sources and demands

with similar energy grades.

2 The proposed design method of heat recovery, based on DTE, is suitable for

the construction of a heat-exchanger network with multiple heat sources and demands.

28
Different HTTDs for different types of heat exchangers could be considered in the

DTE method.

3 The amount of heat recovered in the optimized systems based on the DTE and

pinch methods were the same and 2.08 % higher than that of the conventional one.

Moreover, the energy grade of the remaining heat demand could be reduced by 3.7 °C

in the optimized system based on DTE than the pinch method, which allowed a

reduction in the heat-exchanger area or steam temperature.

Acknowledgments

of
ro
This research was supported by the National Natural Science Foundation of China

-p
(Grant NO. 51638010) and the Innovative Research Group of the National Natural

Science Foundation of China (Grant No. 51521005).

re
lP

References
na

[1] Jiang Y. Heating schemes analysis for medium and large cities in the North China.
ur

Hv & Ac 2000 (in Chinese).

[2] Johnson I, Choate W, Davidson A. Waste heat recovery. Technology and

Jo

opportunities in US industry[R]. BCS, Inc., Laurel, MD (United States), 2008.

[3] Suntivarakorn R, Treedet W. Improvement of boiler's efficiency using heat

recovery and automatic combustion control system [J]. Energy Procedia, 2016,

100: 193-197.

[4] Xu G, Huang S, Yang Y, et al. Techno-economic analysis and optimization of the

heat recovery of utility boiler flue gas. Appl Energy 2013;112(4):907–17

[5] Wang D, Bao A, Kunc W, et al. Coal power plant flue gas waste heat and water

recovery[J]. Appl Energy, 2012, 91(1): 341-348.

29
[6] Li Y, Yan M, Zhang L, et al. Method of flash evaporation and condensation –heat

pump for deep cooling of coal-fired power plant flue gas: latent heat and water

recovery. Appl Energy 2016;172:107–17.

[7] Wang S, Zhang J, Wen Z, et al. Forced convective heat transfer of flue gas on

anticorrosion plane surface with condensation. In: Annals of the assembly for

international heat transfer conference; 2006.

[8] Weber C, Gebhardt B, Fahl U. Market transformation for energy efficient

technologies—success factors and empirical evidence for gas condensing

of
boilers[J]. Energy, 2002, 27(3): 287-315.

[9] Chen Q, Finney K, Li H, et al. Condensing boiler applications in the process

ro
industry[J]. Appl Energy, 2012, 89(1): 30-36.
-p
[10] Wei M, Zhao X, Fu L, et al. Performance study and application of new coal-fired
re
boiler flue gas heat recovery system. Appl Energy 2017;188:121–9.
lP

[11] Chen W, Shi W, Wang B, et al. A deep heat recovery device between flue gas

and supply air of gas-fired boiler by using non-contact total heat exchanger[J].
na

Energy Procedia, 2017, 105: 4976-4982.

ur

[12] Shang S, Li X, Chen W, et al. A total heat recovery system between the flue gas
Jo

and oxidizing air of a gas-fired boiler using a non-contact total heat exchanger [J].

Appl Energy, 2017, 207: 613-623.

[13] Smith R. Chemical process: design and integration[M]. John Wiley & Sons,

2005.

[14] Yu H, Feng X, Wang Y. A new pinch based method for simultaneous selection of

working fluid and operating conditions in an ORC (Organic Rankine Cycle)

recovering waste heat[J]. Energy, 2015, 90: 36-46.

[15] Tan R, Aziz M, Ng D, et al. Pinch analysis-based approach to industrial safety

risk and environmental management[J]. Clean Technol Environ Pol, 2016, 18(7):

2107-2117.

30
[16] Kemp I. Pinch analysis and process integration: a user guide on process

integration for the efficient use of energy[M]. Elsevier, 2011.

[17] Hohmann E., Optimum networks for heat exchange.Ph.D. Thesis,University of

Southern California (1971).

[18] Linnhoff B, Hindmarsh E. The pinch design method for heat exchanger

networks[J]. Chem Eng Sci, 1983, 38(5): 745-763.

[19] Gadalla M. A new graphical method for Pinch Analysis applications: Heat

exchanger network retrofit and energy integration[J]. Energy, 2015, 81: 159-174.

of
[20] Zhang B, Li J, Zhang Z, et al. Simultaneous design of heat exchanger network for

heat integration using hot direct discharges/feeds between process plants[J].

ro
Energy, 2016, 109: 400-411.
-p
[21] Townsend D, Linnhoff B. Surface area targets for heat exchanger networks in
re
process design[C]// IChemE Annual Research Meeting, Bath: 1984.
lP

[22] Saboo A, Morari M, Colberg R. RESHEX: an interactive software package for

the synthesis and analysis of resilient heat-exchanger networks-II: discussion of

na

area targeting and network synthesis algorithms[J]. Comput Chem Eng, 1986,
ur

10(6): 591-599.
Jo

[23] Makwana Y, Shenoy U. A new algorithm for continuous energy targeting and

topology trap determination of heat exchanger networks[M]. Bombay: Ind Inst

Technol, 1993: 178-179.

[24] Trivedi K, O'Neill B, Roach J, et al. A new dual-temperature design method for

the synthesis of heat exchanger networks[J]. Comput Chem Eng, 1989, 13(6):

667-685.

[25] Linnhoff B, Ahmad S. Cost optimum heat exchanger networks I: minimum

energy and capital using simple models for capital cost[J]. Comput Chem Eng,

1990, 14(7): 729-750.

[26] Carlsson A, Franck P. Design better heat exchanger network retrofits[J].Chem

Eng Prog,1993,89(3):87-96.

31
[27] Incropera F, Lavine A, Bergman T, et al. Fundamentals of heat and mass

transfer[M]. Wiley, 2007.

[28] Bargos F, Lamas W, Bilato G. Computational tools and operational research for

optimal design of co-generation systems[J]. Renew Sustain Energy Rev, 2018, 93:

507-516.

[29] Ma Y, Wang Z, Lu J, et al. Techno-economic analysis of a novel hot air

recirculation process for exhaust heat recovery from a 600 MW brown-coal-fired

boiler[J]. Energy, 2018, 152: 348-357.

of
[30] Wu Z. Application of wastewater evaporation concentration technology in

desulphurization wastewater treatment[J]. Huadian Technol, 2012, 34(11):63-66.

ro
[31] Li F, Duanmu L, Fu L, et al. Research and application of flue gas waste heat
-p
recovery in co-generation based on absorption heat-exchange[J]. Procedia Eng,
re
2016, 146: 594-603.
lP

[32] Xu G, Huang S, Yang Y, et al. Techno-economic analysis and optimization of the

heat recovery of utility boiler flue gas[J]. Appl energy, 2013, 112: 907-917.
na

[33] Ye M. Dianchang guolu yuanli ji shebei (V4) [M]. China Electric Power Press,
ur

2016. (in Chinese)

Jo

[34] Ministry of Housing and Urban-Rual Development of the People’s Republic of

China. Design code for heating ventilation and air conditioning of civil

buildings[S]. 2012.10

32
Table 1 Heat source and demands of CHP
Table 2 Heat flow rate difference and accumulated net heat
Table 3 Energy-saving potentials of OHR systems under design conditions
Table 4 Comparison of heat exchanger areas and steam temperatures
Table 5 Areas of heat exchangers in OHRDTE,MHTTD system and OHRDTE system

of
ro
-p
re
lP
na
ur
Jo
 Table 1 Heat source and demands of CHP
Heat source and demands Mass flow rate Inlet temperature Outlet temperature
(kg/s) (°C) (°C)
Flue gas 155.34 200 ~
Supply air 150.8 -10 100
Heating water 1076.55 45 75

Table 2 Heat flow rate difference and accumulated net heat

NO. ∆T CPh-CPc ∆Q Qac

of
(K) (kW/K) (kW) (kW)

ro
1 90 +173.1402 +15582.62 +15582.62
2 25 +21.5862 -p +539.66 +16122.28
re
3 30 -4478.3928 -134351.78 -118229.5
4 55 +21.5862 +1187.24 -115310.87
lP
na

Table 3 Energy saving potentials of OHR systems under design conditions

Texhausted Qrecovery ηrecovery Qpublic Energy grade of
ur

(°C) (MW) (%) (MW) remaining demand

OHR 1 42.15 33.44 8.36 118.23 58.2 °C
Jo

OHR 2 42.15 33.44 8.36 118.23 61.9 °C

CHR 55 25.11 6.28 126.56 60.9 °C

Table 4 Comparison of heat exchanger areas and steam temperatures

HXsteam-water HXgas-water HXgas-air Tsteam
(m2) (m2) (m2) (°C)
OHRDTEa 865 4296.04 19543.57 105
a
OHRDTE
OHRDTEb 943 4296.04 19543.57 101.35
OHR 2 943 4201.54 9259.69 105
CHR 991.63 3132.54 3906.36 105
Table 5 Areas of heat exchangers in OHRDTE,MHTTD system and OHRDTE system
HXsteam-water HXgas-water HXgas-air Tsteam
(m2) (m2) (m2) (°C)
OHRDTE 865 4296.04 19543.57 105
OHRDTE,MHTTD 876.51 7675.46 9259.69 105

of
ro
-p
re
lP
na
ur
Jo
Fig. 1 Theoretical schematics of the discretized thermal energy for the heat sources
and demands
Fig. 2 Energy grade and heat flow rate of the heat sources and demands
Fig. 3 Energy grade matching of the heat sources and demands
Fig. 4 The relationship of accumulation heat sources and heat demands
Fig. 5 Process of simplification method
Fig. 6 Energy flow chart of the CHP plant
Fig. 7 The conventional condensing heat recovery system

of
Fig. 8 Optimization processes for OHRDTE system

ro
Fig. 9 The temperature interval of heat source and demands in pinch method
Fig. 10 Final system of OHRpinch system for implementation
-p
Fig. 11 Heat source and demand in each energy grade with different HTTD
re
Fig. 12 Schematics of OHR3 system with different HTTDs
lP
na
ur
Jo
Fig. 1 Theoretical schematics of the discretized thermal energy for the heat sources
and demands

of
ro
-p
re
lP
na
ur

Fig. 2 Energy grade and heat flow rate of the heat sources and demands
Jo
 of
Fig. 3 Energy grade matching of the heat sources and demands

ro
-p
re
lP
na
ur
Jo

Fig. 4 The relationship of accumulation heat sources and heat demands

(a) The expression of the heat exchanger

(b) Actual heat exchanger

Fig. 5 Process of simplification method
 of
Fig. 6 Energy flow chart of the CHP plant

ro
-p
re
lP
na
ur
Jo

Fig. 7 The conventional condensing heat recovery system

 (a) Division of heat grade range

of
ro
-p
re
lP
na
ur

(b) Heat source and demand in each energy grade

Jo

(c) Theoretical system of the optimized heat recovery system

 of
(d) Final system for implementation

ro
Fig. 8 Optimization processes for OHRDTE system
-p
re
lP
na
ur
Jo

Fig. 9 The temperature interval of heat source and demands in pinch method

Fig. 10 Final system of OHRpinch for implementation

 of
Fig. 11 Heat source and demand in each energy grade with different HTTD

ro
-p
re
lP
na
ur
Jo

Fig. 12 Schematics of OHR3 system with different HTTDs

Highlights

• Method of discretizing thermal energy for heat sources and demands is proposed.
• An optimization method to match heat sources and demands is established.
• Two heat recorvery systems are developed based on the proposed method.
• The proposed method has the same heat recorvery with pinch method.
• Energy grade of remaining heat demand is lower than that with pinch method.

of
ro
-p
re
lP
na
ur
Jo
Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

of
ro
-p
re
lP
na
ur
Jo