Applied Thermal Engineering 181 (2020) 115996

Contents lists available at ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

A comprehensive energy efficiency study of segmented annular

thermoelectric generator; thermal, exergetic and economic analysis
Man-Wen Tian a, Leonardus W.W. Mihardjo b, Hazim Moria c, Soheil Asaadi d, *,
Hamed Sadighi Dizaji d, Shahram Khalilarya d, Phong Thanh Nguyen e, *
a
Institute of Applied Economics, Yango University, Fuzhou 350015, China
b
Business School, Bina Nusantara University, Jalan Hang Lekir I no. 6, Senayan, Jakarta 10270, Indonesia
c
Department of Mechanical Engineering Technology, Yanbu Industrial College, Yanbu Al-Sinaiyah City 41912, Saudi Arabia
d
Department of Mechanical Engineering, Urmia University, Urmia, Iran
e
Department of Project Management, Ho Chi Minh City Open University, Ho Chi Minh City, Viet Nam

A R T I C L E I N F O A B S T R A C T

Keywords: Annular thermoelectric generator (ATEG) has not been commercialized yet and many studies are required to
Thermoelectric identify its characteristics from different viewpoints. Most recently, segmented annular thermoelectric was
Generator proposed as an efficient structure for annular thermoelectric in which each leg of thermoelectric (n-type and p-
Exergy
type) is made of two different materials termed low temperature material and high temperature material.
Economy
Although a few of investigations have reported thermal behavior of segmented annular thermoelectric, no
exergetic and economic analysis have been performed yet on segmented annular thermoelectric generators.
Exergetic and economic evaluations are significant criteria on marketing area of any industrial product. Hence, in
this study, a comparison between segmented and non-segmented annular thermoelectric is first presented and
then the impacts of the main parameters of segmented annular thermoelectric generator on energy, exergy and
economic factors (including output power, conversion efficiency, exergy efficiency and cost per unit of output
power) are comprehensively investigated and presented via validated 3-D numerical simulation method. The
results showed that the segmented thermoelectric is more efficient than the non-segmented thermoelectric from
the viewpoint of all said criteria. A peak point was observed for the curve behavior of most geometric charac­
teristics of the segmented annular thermoelectric generator which is remarkable and discussed in this paper.

including exploring and employment of more efficient materials [5–7]

(i.e. high electrical conductivity and low thermal conductivity),
geometrical optimization [8–11], application of transient and pulsed
1. Introduction
inputs [12–15], and use of segmented [16] and two-stage [17–19]
structures of thermoelectric legs. As the progress of exploring of the
Although the fossil fuels are the main source of energy, the
more efficient materials is not practically sufficient yet, working on
increment of their negative aspects have encouraged the scientists to
design and fabrication process of existent materials has become
propose the alternation of them [1,2]. These negative features include
precious, valuable and popular [15]. Segmented leg was introduced as
un-renewable nature, air pollution, increment of energy demand and
an efficiency improvement technique. Main studies in this area are
climate change [3]. Entrepreneurs and investors have not been
summarized in the following.
convinced yet to widespread consider the thermoelectric device as an
Shen et al. [20] developed a theoretical solution for the one
efficient renewable source of energy. The most significant reason lies
dimensional model of an annular thermoelectric couple (ATEC). They
under lower performance of thermoelectric devices compared to the
derived the fundamental formulas of annular thermoelectric in the
other existence energy sources. Hence, scientists have been looking
steady-state condition. Kaushik and Manikandan [21] presented an
for novel techniques to enhance the thermal efficiency of various
analytical solution for the energy performance of annular thermo­
systems that deal with the rejection of heat [4]. The most important
electric generator (ATEG) considering the influence of the Thomson
efforts in this regard are classified into some main categories

* Corresponding authors.
E-mail addresses: [email protected] (S. Asaadi), [email protected] (P.T. Nguyen).

https://doi.org/10.1016/j.applthermaleng.2020.115996
Received 18 April 2020; Received in revised form 15 July 2020; Accepted 31 August 2020
Available online 8 September 2020
1359-4311/© 2020 Elsevier Ltd. All rights reserved.
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

Nomenclature Greek letters

( ) β, φ angle, (degree)
A cross-sectional area of thermoelectric leg, m2 η conversion efficiency, energy efficiency
( )
E electric field intensity vector, Vm− 1 Δ difference
Ex exergy, (W) ρ density, (kgm− 3 )
H height, (m) σ electrical conductivity, (Sm)
( )
I electric current, (A) τ Thomson coefficient, VK− 1
( )
J electric current density vector, Am− 2
( ) Subscript
k thermal conductivity, Wm− 2 K− 1
0 environment
P power output, (W)
c cold side of ATEG
Q heat, (W)
( ) cr ceramic
q˝ heat flux, Wm− 2 e exiting
R electrical resistance, (Ω) h hot side of ATEG
r radius, (m) i entering
( )
S Seebeck coefficient, VK− 1 ic interconnector
T temperature, (K) in input
W thickness, (m) L Lower segment
V Electrical voltage, (V) Load load
Z capital investment cost m mean temperature
ZT Thermoelectric figure of merit n n-type
oc open-circuit
Abbreviations
out output
ATEC annular thermoelectric couple
p p-type
ATEG annular thermoelectric generator
t thermocouple
STEG segmented thermoelectric generator
sr semiconductor
SATEG segmented annular thermoelectric generator
U upper segment

effect. They concluded that annular thermoelectric is more useful in a of conversion of a flexible carbon nanotube p-type into an n-type ma­
variety of applications such as recovering waste heat from solar en­ terial with polyethylenimine concept. Yin et al. [35] experimentally
ergy and exhaust gas due to its annular shape. Fan and Gao [22] optimized the integrated photovoltaic and thermoelectric solar gener­
numerically studied the mechanical and thermodynamic performance ator. They verified the feasibility of this system and indicated that the
of an annular thermoelectric generator. According to their results, electrical load resistance of thermoelectric should be smaller than the its
increasing the length of thermoelectric legs can improve the me­ internal resistance.
chanical reliability of the system, but decreases the efficiency of the Segmented annular thermoelectric is comprehensively studied in this
ATEG. Shen et al. [23] studied the performance of ATEC under con­ paper from the energy, exergy and economy viewpoints. Although a few
stant heat flux and constant temperature. The thermoelectric perfor­ of investigations [30,36] have reported thermal behavior of segmented
mance of ATEG under pulsed heat condition with a different type of annular thermoelectric, no exergetic and economic analysis have been
heat input functions has been studied numerically by Asaadi et al. performed before on segmented annular thermoelectric generators.
[24]. They reported that the pulsed heat input can improve the per­ Economic evaluation is a significant criterion in marketing area of any
formance of ATEG devices. industrial product. At the first stage of this paper, a comparison is pro­
As mentioned above, using segmented design is another approach vided between simple annular thermoelectric and segmented annular
that can lead to enhancement in the performance of thermoelectric de­ thermoelectric. At the second step, the impacts of the most important
vices. As no single material has high efficiency in all temperature ranges, parameters of segmented annular thermoelectric including height
employment of two embedded materials one of which is efficient in high ration, angle ratio and load ration on energetic, exergetic and economic
temperatures and the other is efficient in low temperatures will be a factors including output power, energy efficiency, exergy efficiency and
solution [25–29] termed segmented leg. The researches on the cost per unit of output power) are investigated and reported via vali­
segmented annular thermoelectric devices are so limited and requires dated 3D numerical simulation.
more investigations. Shen et al. [30] performed a theoretical analysis on
the segmented annular thermoelectric generator. Results proved that the
segmented design has better performance than simply non-segmented 2. Materials and geometries
ATEG. Shittu et al. [31] reported a 21.7% and 82.9% increase of effi­
ciency for SATEG compared to ATEG. Scherrer et al. [32] studied As shown in graphical abstract, the investigation section of this study
numerically on a flat shape solar thermoelectric generator based on can be divided into four main categories including “comparison between
skutterudited material using optimal control theory. They tried to pre­ non-segmented and segmented thermoelectric”, “the effect of height
sent optimum geometry and weight for solar thermoelectric generators. ratio”, “the effect of angle ratio” and “the effect of load ratio” on ener­
Dou et al. [33] performed an investigation of bismuth telluride nano­ getic, exergetic and economic factors. In each category, some parame­
tubes applied in solar cell. Bharti et al. [34] tried to clarify the possibility ters are variant and some others are constant. All required dimensions

2
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

Fig. 1. Dimensions, materials and properties of tubular thermoelectric used in this study.

[ ]/
and materials used in this paper are shown in Fig. 1. The non-segmented (1 + ZT)0.5 − 1 ST). The goal is to select a high figure of merit ma­
ATEG consisted of n-type and p-type thermoelectric legs which are
terials that have similar compatibility factors [38].
connected by two curved blocks of copper interconnectors. Alumina
electrical insulator ceramics are attached to the inner and outer sides of
3. Mathematical modelling and equations
ATEG [37]. However, in the case of SATEG, the thermoelectric legs are
composed of high-temperature segment and low-temperature segment.
The mathematical modelling process can be divided into four main
The inner side of ATEG is considered as a heat source while the outer
categories including general governing equations, energy parameters,
side acts as the heat sink. In is noted that, the materials are generally
exergetic parameters and economic analysis which summarized in the
selected based on the figure of merit (ZT), and compatibility factor (ṡ =

3
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

following with more explanation for economic modelling. However,

some considerations are required for segmented thermoelectric which
will be explained.

It is noted that, in the ATEG, there are two kinds of irreversibilities,

internal and external. The material properties and the contact resistance
between thermocouples and metal are the cause of internal irrevers­
ibilities, while external irreversibilities are due to irreversible heat
transfer [39,40]. The effect of contact resistances is eliminated in this
study, and material properties are the sole cause of internal irrevers­
ibilities. For thermoelectric generators, the exergy output of the system
is equal to power output which is 100% exergy.
In a steady-state operating condition, for any system, there may be
some material streams in and out of the system as well as heat and work
interactions with the surroundings. The stream of materials and energy
In the case of SATEG, the thermoelectric legs are composed of two to the systems is associated with exergy transfer [41] and the irrevers­
materials in each leg. The voltage of SATEG is defined by the sum of the ibilities within the system cause exergy destruction. The specific exergy
voltage at segments, which is described with the help of interface tem­ costing (SPECO) method proposed by Lazzaretto and Tsatsaronis [42] is
peratures at materials interfaces [30]. In Eq. (13), p, n, L, U, m indexes used to carry out the economic analysis of segmented annular thermo­
correspond to p-type, n-type, lower segment, upper segment, and electric. To apply the SPECO method in ATEG, the few following steps
interface, respectively. are considered:

(i) Energy and exergy are derived for each of the supposed points in
ATEG.
Exergy equations are shown as below, (ii) The definition of fuel and product is presented for each part of
ATEG.

4
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

Table 1
Definition of fuel and product for non-segmented and segmented ATEG. Zt = Zt ’’’ Vt + Zt ’’ At (22)

Thermoelectric Part of thermoelectric Fuel Product

Zcr = Zcr ’’ Acr (23)
ATEG Hot side ceramic EXQH EX1
Thermocouples EX1 EX2 + POut1 Z = Zt + Zcr (24)
Cold side ceramic EX2 EXQC
In the above equations, the terms are volumetric cost Zt ’’’ ($/mm3 )
SATEG Hot side ceramic EXQH1 EX1
Lower segment of legs EX1 EXM1
and areal cost Zt ’’ ($/mm2 ). In the case of the segmented annular ther­
Upper segment of legs EXM1 EX2 + POut2
moelectric generators the capital cost of thermocouples is the sum of the
Cold side ceramic EX2 EXQC
costs in high-temperature and low-temperature segments as:

Zt = ZtL ’’’ VtL + ZtL ’’ AtL +ZtU ’’’ VtU + ZtU ’’ AtU (25)

(iii) Cost balances are formulated for each component of ATEG The capital cost of materials is defined in the form of cost rate ($/s) as
separately. [43,44]:

Ż k = ZK .CRF.ϕ/(N × 3600) (26)

The fuel and product definition are presented for each part of simply
non-segmented and segmented ATEG in Table 1.
i(1 + i)m
For any component of the ATEG system, the total cost of the exiting CRF = (27)
(1 + i)m − 1
exergy stream equals the total expenditure to obtain them: the capital
and other costs plus the cost of the entering exergy streams [41]. This
where N is the operating hours of ATEG during a year (7500 h) [45],
can be described as:
CRF is capital recovery factor, ϕ is the maintenance factor (ϕ = 1 due to
∑ ∑
Ċe,k + Ċw,k = Ċq,k + Ċi,k + Ż k (20) maintenance condition [43]), m is the life period of the system (20
e i years) and i is interest rate (10%) [44]. The volumetric and areal costs of
employed materials are listed in Table 2. The cost rate balance equations
where for each part of non-segmented ATEG and SATEG are presented in
Table 3.
˙j
Ċj = ċj EX (21)

where i and e correspond to exergy streams that transfer in and out of kth 4. Assumptions, boundary conditions and simulation process
component. The capital cost for ceramics and thermocouples in ATEG is
calculated as: [43]: In numerical modeling and simulations, the following assumptions
are considered:

• Heat transfer from the side surfaces of ATEG to the surroundings and
Table 2 contact thermal and electrical resistances are ignored.
Areal and volumetric costs of TE materials [46]. • The influence of the Thomson effect is considered.
Material Z’’’ ($/mm3 ) Z’’ ($/mm2 )

4 4
Bi2 Te3 8.89565 × 10− 1.6823 × 10−
Bi0.52 Sb1.48 Te3 8.65743 × 10− 4
1.6823 × 10− 4

− 4 − 4
CeFe4 Sb12 2.62299 × 10 1.6823 × 10
4 4
Yb0.2 In0.2 Co4 Sb12 1.93688 × 10− 1.6823 × 10−
Alumina Ceramic 8.6251 × 10− 7

Table 3
Cost balance equations for different parts of non-segmented ATEG and STAEG.
Thermoelectric Part of thermoelectric Exergy cost rate balance equation

Hot side ceramic ĊQH + ŻCr = Ċ1

ATEG Thermocouples Ċ1 + Żt = Ċ2 + Ċp
Cold side ceramic Ċ2 + ŻCr = ĊQC
Hot side ceramic ĊQH1 + ŻCr = Ċ1
STAEG Lower segment of legs Ċ1 + ŻM1 = ĊM1
Upper segment of legs ĊM1 + ŻM2 = Ċ2 + Ċp2 Fig. 2. Comparison of the present simulation results for non-segmented ATEG
Cold side ceramic Ċ2 + ŻCr = ĊQC with the analytical results by Gao Shen et al. [20].

Table 4
Results of grid independence test.
Non-segmented ATEG STAEG

Number of grids Voltage (V) Pmax (W) Number of grids Voltage (V) Pmax (W)

3219 0.02282100 0.01658685 3212 0.03238900 0.02401283

5452 0.02281600 0.01657959 5686 0.03238900 0.02401283
11,215 0.02281300 0.01657523 11,374 0.03238200 0.02400297
38,779 0.02281100 0.01657250 38,874 0.03238000 0.02399994

5
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

Fig. 3. A general view of non-segmented and segmented annular thermoelectric generator.

Fig. 4. Comparison of performance and cost between non-segmented and segmented ATEGs (a) maximum output power (b) conversion efficiency (c) exergy effi­
ciency (d) cost per unit of output power.

To simulate the thermoelectric model, the COMSOL Multi-physics The electrical ground condition V = 0 is applied to the outer left
5.4 is used. The three-dimensional equations of the system of the boundary of the copper interconnector. Moreover, the exergoeconomic
annular thermoelectric generator are solved using the finite element study is carried out by putting the numerically calculated results of the
method. Constant temperatures Th and Tc are employed at the heat energy analysis into the exergy and cost balance equations.
source (lower side) and heat sink (upper side) boundaries, respectively.

6
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

5. Grid independency and validation of numerical results

The discretization of the simulated domain is applied based on the

tetrahedral unstructured grid system. The grid independence study is
performed for simply non-segmented and segmented annular thermo­
electric generators. The variation of voltage and maximum output power
for different grid systems for non-segmented and segmented ATEGs is
shown in Table 4. As shown in Table 4, the results of all grid systems are
similar, and the grid system with the lowest number of elements can be
applied in numerical simulation process to reduce the processing time.
Moreover, the validation of simulation is conducted with the analytical
results of Gao Shen et al. [20] for non-segmented ATEG. The variation of
voltage and maximum power output versus heat source temperature are Fig. 5. A general view of segmented annular thermoelectric generator in height
shown in Fig. 2 for the present study in comparison with analytical ratio of a) 1 and b) 10.
expression. Almost no difference is shown between the numerical results
and analytical solution. The maximum error of 0.85% and 1.71% be­
can produce more power for the lesser amount of absorbed heat in this
tween simulation results and the analytical expression for voltage and
range of temperatures. Moreover, it can be seen that the amount of
output power were found.
improvement in conversion efficiency is significant compared to the
increment of output power. From the viewpoint of exergy efficiency,
6. Results and discussion
very similar behavior can be observed, and for a wide range of heat
source temperatures the exergy efficiency increase by a good amount.
6.1. Comparison between non-segmented and segmented annular
For example, in Th = 598.15K, the exergy efficiency of SATEG is 37.2%
thermoelectric
higher than the exergy efficiency of non-segmented ATEG composed of
low-temperature materials and is 41.3% higher than non-segmented
A general view of non-segmented and segmented tubular thermo­
ATEG composed high-temperature materials [40,43]. Also, by
electric material are presented in Fig. 3. A comparison between non-
increasing the temperature, output power, and conversion and exergy
segmented annular thermoelectric and segmented thermoelectric in
efficiency increases.
terms of Pmax, conversion efficiency, exergetic efficiency, unit cost of
Fig. 4(d) represents the variation in the cost per unit of output power
output power and the results are illustrated in Fig. 4. The variant
for non-segmented and segmented ATEGs. It can be seen that by
parameter is hot source temperature and all other parameters are as
increasing temperature, the cost of output power decreases. The exer­
presented in Fig. 1 before.
goeconomic study reveals that the SATEG performs more economical
Based on Fig. 4, for the temperature between 398 K and 598 K, the
than the non-segmented ATEGs. As described above, segmented ther­
segmented thermoelectric provides averagely 27% (8% at 398 K and
moelectric comprises of two materials (lower temperature material and
45% at 598 K) higher output power compared to the Bismuth tellurides.
high temperature material). The segmented ATEG (with material A and
Moreover, in the mentioned temperature range, segmented thermo­
B), is economical than the non-segmented ATEG made of single material
electric provides averagely 19% (33% at 398 K and 6% at 398 K) higher
A or B. Moreover, the main portion of the required cost is related to the
output power compared to the Skutterudites. For the temperatures
cost of exergy stream and material cost does not play a key role in
higher than 598 K, Skutterudites (high temperature material) provides
required cost of thermoelectric generator. That is why the SATEG should
higher output power. The difference between Skutterudites and
be economical compared to the non-segment thermoelectric for other
segmented thermoelectric is further increased by more increment of
materials as well. Nonetheless, other materials may act economically in
source temperature. For the temperature lower than 398 K, the output
other range of temperature as their characteristics varies with temper­
power is almost the same (however, around 2% more power is shown for
ature variation. Referring to the results of the study on the two-stage
Bismuth tellurides compared to the others in numeric results).
annular thermoelectric generator by Asaadi et al. [43], the present
One side of any thermoelectric generator is in contact with the high
investigation proves that the segmented design leads to lower costs
temperature and the other side is in contact with low (cold) tempera­
compared to two-stage design studied by [43]. While both segmented
ture. However, any leg material of thermoelectric is able to act effec­
and two-stage design improve the energy and exergy performance of
tively only in cold range temperatures (such as Bismuth tellurides) or
ATEG devices, segmented design leads to the increment of energy and
hot range temperatures (such as Skutterudites). As in segmented ther­
exergy efficiency with lesser costs.
moelectric, the hot side of thermoelectric is made of high temperature
In the following, the effect of geometric parameters, including the
material (Skutterudites) and the cold side of thermoelectric is made of
height ratio of segments and angle ratio of legs and load ratio on the
low temperature material (Bismuth tellurides), its efficiency is higher
different performance criterion is studied. Output power, conversion
than the non-segmented thermoelectric in which the whole leg is made
and exergy efficiency of segmented annular thermoelectric generators
of only one material. In other words, if the whole leg is made of high
are investigated. Besides the energy and exergy study, the exer­
temperature material, the cold side will not work effectively and if the
goeconomic analyses are carried out to further reveal the cost per unit of
whole leg is made of low temperature material, the hot side will not
generated power. It is assumed that the heat is supplied to ATEGs by a
work effectively. That is why segmented thermoelectric provides higher
diesel engine that operates in a heat and power cogeneration system
output power
developed by Mohammadkhani [45]. The exhausted gas with a
Conversion (energy) and exergy efficiency of ATEG devices depends
maximum temperature of 908.05 K provides the required input energy,
on on the amount of absorbed heat as well as the output power
which is harvested by ATEG units. The cost per unit of input heat is
[47,48,49]. Taking this into account, a device can produce more power
derived by an exergoeconomic study equal to 5.99 ($/GJ). Finally, the
for a specific range of temperatures while having less efficiency. As
influence of the load ratio RLoad /R on the efficiency and cost perfor­
shown in Fig. 4(b) the SATEG has higher conversion efficiency for the
mance of SATEG is presented. It is worth mentioning that in all of the
heat source temperature above 448.15 K, which means that the SATEG
simulations, the thermal and flow resistance between exhausted gas flow

7
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

Fig. 6. Effect of HU/HL on the performance of SATEG (a) maximum output power (b) conversion efficiency (c) exergy efficiency (d) cost per unit of output power.

and 10 for different hot side temperatures. All other constant parameters
was presented in Fig. 1. A view of segmented annular thermoelectric in
the minimum and maximum value of height ratio is shown in Fig. 5. The
results of the comparison are presented in Fig. 6.
Referring to the Fig. 6, the source temperature is a determinant factor
in the thermal and electrical behavior of SATEG. Skutterudites performs
better at high temperatures, and bismuth tellurides have better perfor­
mance at low temperatures. The volume of each material in the SATEG is
defined by height ratio, and for different source temperatures, the peak
value of performance occurs at different height ratios. As can be seen in
Fig. 6(a) while temperature grows, the maximum output power is ob­
tained at lower height ratios. This outcome is not surprising since the
skutterudites are the materials with better performance at higher tem­
Fig. 7. A general view of segmented annular thermoelectric generator in height peratures, and in a lower height ratio, a bigger percentage of legs are
ratio of a) 1 and b) 10. composed of better performing materials. In all operating temperatures,
similar results are obtained for conversion efficiency. With the incre­
and ATEG units are ignored, reflecting that the heat source temperature ment of temperature, the peak of conversion efficiency occurs at lower
is exactly the same as exhaust gas temperature [50–51]. height ratios. For example, in Th = 398.15, 598.15 and 798.15K the
maximum conversion efficiencies are 0.0419, 0.0825 and 0.1191 which
are resulted in HHUL = 10, 0.7 and 0.3 respectively. A similar trend cannot
6.2. Effect of height ratio HU /HL
be observed for the exergy efficiency of SATEG, as depicted in Fig. 6(c).
The influence of the height ratio Hu/HL on different performance As described in the modeling section, the exergy efficiency of the SATEG
criteria including power output, energy efficiency, exergy efficiency, can be obtained by dividing the conversion efficiency with the Carnot
and cost per unit of output power is studied in this section for segmented factor, (1 − T0 /Th ). The Carnot factor and the conversion efficiency are
annular thermoelectric generator. The height ratio was varied between 1 two determining factors that are both affected by the temperature of the

8
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

Fig. 8. Effect of (φn/φp) on the performance of SATEG (a) maximum output power (b) conversion efficiency (c) exergy efficiency (d) cost per unit of output power.

heat source and sink as well as the height ratio. Increasing heat source thermoelectric. The angle ratio was varied between 1 and 10 for
temperature Th enhances the conversion efficiency as mentioned, but different hot side temperatures. All other constant parameters was
the Carnot factor grows as well. Taking this into account, between these presented in Fig. 1. A view of segmented annular thermoelectric in the
two conflicting factors, the outcome is determined by the dominant minimum and maximum value of angle ratio is shown in Fig. 7. The
factor. For a specific range of height ratios with the increase of source results of the comparison are presented in Fig. 8.
temperature, the increase of the Carnot factor is more than the As can be seen in Fig. 8(a), for all of the tested temperatures, with the
enhancement of conversion efficiency, and as a result, the overall exergy increment of angle ratio between 0.1 and 1, the output power ap­
efficiency decreases. proaches the peak value and then begins to drop. This finding is in good
Fig. 6(d) depicts the changes in the cost per unit of output power agreement with the results of Asaadi et al. [43] for the two-stage annular
versus height ratio for different temperatures. According to the thermoelectric generator. The output power decreases faster when the
modeling section and equations of cost balance and based on exer­ angle ratio grows to values more than 1. This is because the n-type
goeconomic analysis, the cost per unit of power output is dependent on materials have higher electrical resistivity values, and when the mate­
various parameters including, areal and volumetric costs of materials, rials with higher electrical resistivity are applied to the leg with the
exergy streams costs, and output power. A material with lower areal and bigger angle, the total electrical resistance of SATEG increases which
volumetric costs is always favorable. It should be considered that the leads to lower output power [30]. Conversion efficiency and exergy
exergy streams make up a greater share of costs compared to material efficiency of SATEG change with the variation of the angle ratio in a very
costs. Higher output power values lead to the reduction of cost, and since similar trend to output power, referring to Fig. 8(b) and (c). As can be
high heat source temperature increases output power, the lesser cost can seen, an optimum angle ratio equal to 1 results the peak conversion and
be expected. exergy efficiency. For example in Th = 398.15and 598.15K the obtained
conversion efficiency and exergy efficiency in φn /φp = 0.9, 1 and 2 are
(0.03393, 0.1356), (0.03399, 0.1358) and (0.3081, 0.1231)
6.3. Effect of angle ratio φn /φp
respectively.
The unit cost per unit of output power for a wide range of operating
The influence of the angle ratio (φn /φp )on different performance
heat source temperatures versus the variation of angle ratio is illustrated
criteria including power output, energy efficiency, exergy efficiency,
in Fig. 8(d). According to the figure, the unit cost per unit of output
and cost per unit of output power is studied in this section for segmented

9
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

Fig. 9. Effect of (RLoad/R) on the performance of SATEG (a) maximum output power (b) energy efficiency (c) exergy efficiency (d) cost per unit of output power.

power decreases with the increment of angle ratio between 0.1 and 1 proposed a range between 1 and 1.3 to get the best possible perfor­
and then increases with the further increase of angle ratio. It is discussed mance. Referring to Fig. 7(c) the trend of exergy efficiency versus RLoad /
that the output power and exergy streams largely influence the cost per R is similar to energy efficiency and the peak value of this quantity
unit of output power. It can be expected that the angle ratio with the cannot be obtained in the same RLoad /R for all of the applied heat source
highest output power and efficiency leads to the lowest cost. All ob­ temperatures. On the other hand, it is clear that the improvement of the
tained results confirm that the angle ratios in the vicinity of 1 are the energy efficiency and exergy efficiency among the mentioned range are
optimum choice to improve the efficiency of the SATEG device while relatively small and can be ignored to simplify the control and lower the
keeping cost low. associated costs. Fig. 9(d) reveals that as RLoad /R increases, the cost per
unit of output power decreases in the first stage and then is increases. It
6.4. Effect of load ratio RLoad /R can be seen that the RLoad /Rcorresponding to the least cost is equal to 1
for heat source temperatures lower than 598.15 K, and for higher tem­
The changes in output power, energy efficiency, exergy efficiency, peratures, it increases to the 1.5 influenced by exergy performance of the
and cost per unit of output power are plotted in Fig. 9. All other constant system. It is shown that in 1 < RLoad /R < 1.5 the cost of power can be
parameters was presented in Fig. 1. As can be seen with increasing the minimized depending on the source temperature.
ratio of RLoad /R the output power increases up to a maximum value and
declines afterward. The peak of power can be obtained in RLoad /R = 1, 7. Conclusion
which is expected as explained in the modeling section. The energy and
exergy efficiency both share a similar trend with output power. In this paper segmented annular thermoelectric generator (SATEG)
Although the output power and energy efficiency share a similar trend, was first compared with two non-segmented thermoelectric including
the maximum output power and energy efficiency cannot be obtained Skutterudites (high temperature material) and Bismuth tellurides (low
simultaneously for all source temperatures. While power can be maxi­ temperature material). SATEG provided higher power in the tempera­
mized whenRLoad /R = 1, the peak energy efficiency happens at RLoad / ture range of 398 K and 598 K. However, for the temperatures higher
R = 1.5 for heat source temperatures higher than 598.15 K. Taking all than this range, Shutterudites generates more power compared to the
these to account an optimal range of RLoad /R about 1–1.5 can be STAG. In the next step, the effects of key parameters of SATEG including
considered. A similar result is reported by Shen et al. [30]. They height ratio, angle ratio and load ration are varied and their effects on

10
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

thermal, exergetic and economic factors are investigated. The cost per [13] L.M. Shen, F. Xiao, H.X. Chen, S.W. Wang, Numerical and experimental analysis of
transient supercooling effect of voltage pulse on thermoelectric element, Int J
unit of output power for SATEG was found lower than the non-
Refrig 35 (2012) 1156–1165, https://doi.org/10.1016/j.ijrefrig.2012.02.004.
segmented thermoelectric. The main factor of cost per unit of output [14] M. Ma, J. Yu, J. Chen, An investigation on thermoelectric coolers operated with
power is the cost of exergy stream and material cost does not have sig­ continuous current pulses, Energy Convers Manage 98 (2015) 275–281, https://
nificant impact. SATEG also provided higher energy and exergy effi­ doi.org/10.1016/j.enconman.2015.03.105.
[15] A.S. Al-Merbati, B.S. Yilbas, A.Z. Sahin, A model study for cyclic thermal loading
ciency in desired range of source temperature which shows the and thermal performance of a thermoelectric generator, Int J Energy Res 38 (2014)
suitability of the SATEG compared to the non-segmented generator. The 1351–1360, https://doi.org/10.1002/er.3152.
curve behavior of output power against the tested geometric or thermal/ [16] T. Ming, Y. Wu, C. Peng, Y. Tao, Thermal analysis on a segmented thermoelectric
generator, Energy 80 (2015) 388–399, https://doi.org/10.1016/j.
electric parameters showed extremum points which are remarkable. In energy.2014.11.080.
all of the tested heat source temperatures, as angle ratio increases, the [17] R. Arora, S.C. Kaushik, R. Arora, Thermodynamic modeling and multi-objective
output power approaches the peak value and declines afterward. The optimization of two stage thermoelectric generator in electrically series and
parallel configuration, Appl Therm Eng 103 (2016) 1312–1323, https://doi.org/
cost of power can be minimized between the load ration of 1 and 1.5 10.1016/j.applthermaleng.2016.05.009.
depending on source temperature. As SATEG was also found more effi­ [18] Z. Liu, S. Zhu, Y. Ge, F. Shan, L. Zeng, W. Liu, Geometry optimization of two-stage
cient compared to the two-stage thermoelectric generators investigated thermoelectric generators using simplified conjugate-gradient method, Appl
Energy 190 (2017) 540–552, https://doi.org/10.1016/j.apenergy.2017.01.002.
in previous researches, it is recommended as a promising performance [19] J. Chen, J. Yu, M. Ma, Theoretical study on an integrated two-stage cascaded
enhancement technique for annular thermoelectric generators. thermoelectric module operating with dual power sources, Energy Convers Manage
98 (2015) 28–33, https://doi.org/10.1016/j.enconman.2015.03.090.
[20] Z.-G. Shen, S.-Y. Wu, L. Xiao, Theoretical analysis on the performance of annular
Declaration of Competing Interest thermoelectric couple, Energy Convers Manage 89 (2015) 244–250, https://doi.
org/10.1016/j.enconman.2014.09.071.
[21] S. Manikandan, S.C. Kaushik, The influence of Thomson effect in the performance
The authors declare that they have no known competing financial optimization of a two stage thermoelectric generator, Energy 100 (2016) 227–237,
interests or personal relationships that could have appeared to influence https://doi.org/10.1016/j.energy.2016.01.092.
the work reported in this paper. [22] S. Fan, Y. Gao, Numerical simulation on thermoelectric and mechanical
performance of annular thermoelectric generator, Energy 150 (2018) 38–48,
https://doi.org/10.1016/j.energy.2018.02.124.
Acknowledgements [23] Z.-G. Shen, S.-Y. Wu, L. Xiao, Assessment of the performance of annular
thermoelectric couples under constant heat flux condition, Energy Convers Manage
150 (2017) 704–713, https://doi.org/10.1016/j.enconman.2017.08.058.
This work was financially supported by the Key project of the Na­ [24] S. Asaadi, S. Khalilarya, S. Jafarmadar, Numerical study on the thermal and
tional Social Science Fundation of the year 2018(18AJY013); the 2017 electrical performance of an annular thermoelectric generator under pulsed heat
National Social Science foundation project (17CJY072); the 2018 power with different types of input functions, Energy Convers Manage 167 (2018),
https://doi.org/10.1016/j.enconman.2018.04.085.
planning project of philosophy and social science of zhejiang province [25] C. Hadjistassou, E. Kyriakides, J. Georgiou, Designing high efficiency segmented
(18NDJC086YB); the 2018 Fujian Social Science Planning Project thermoelectric generators, Energy Convers Manage 66 (2013) 165–172, https://
(FJ2018B067);The Planning Fund Project of Humanities and Social doi.org/10.1016/j.enconman.2012.07.030.
[26] H.S. Kim, K. Kikuchi, T. Itoh, T. Iida, M. Taya, Design of segmented thermoelectric
Sciences Research of the Ministry of Education in 2019 (19YJA790102).
generator based on cost-effective and light-weight thermoelectric alloys, Mater Sci
Eng B 185 (2014) 45–52, https://doi.org/10.1016/j.mseb.2014.02.005.
References [27] G. Zhang, L. Fan, Z. Niu, K. Jiao, H. Diao, Q. Du, et al., A comprehensive design
method for segmented thermoelectric generator, Energy Convers Manage 106
(2015) 510–519, https://doi.org/10.1016/j.enconman.2015.09.068.
[1] S.R. Yan, T. Gholami, O. Amiri, M. Salavati-Niasari, S. Seifi, M. Amiri, M. Sabet, L.
[28] H. Ali, B.S. Yilbas, F.A. Al-Sulaiman, Segmented thermoelectric generator:
K. Foong, Effect of adding TiO2, SiO2 and graphene on of electrochemical
Influence of pin shape configuration on the device performance, Energy 111 (2016)
hydrogen storage performance and coulombic efficiency of CoAl2O4 spinel,
439–452, https://doi.org/10.1016/j.energy.2016.06.003.
J. Alloys Compounds. 5 (828) (2020 Jul), 154353.
[29] H. Ali, B.S. Yilbas, Innovative design of a thermoelectric generator of extended legs
[2] S.R. Yan, M.A. Fazilati, N. Samani, H. Ghasemi, D. Toghraie, Q. Nguyen,
with tapering and segmented pin configuration: Thermal performance analysis,
A. Karimipour, Energy efficiency optimization of the waste heat recovery system
Appl Therm Eng 123 (2017) 74–91, https://doi.org/10.1016/j.
with embedded phase change materials in greenhouses: A thermo-economic-
applthermaleng.2017.05.066.
environmental study, J. Storage Mater. 1 (30) (2020 Aug), 101445.
[30] Z.-G. Shen, X. Liu, S. Chen, S.-Y. Wu, L. Xiao, Z.-X. Chen, Theoretical analysis on a
[3] M.W. Tian, H.C. Yuen, S.R. Yan, W.L. Huang, The multiple selections of fostering
segmented annular thermoelectric generator, Energy 157 (2018) 297–313, https://
applications of hydrogen energy by integrating economic and industrial evaluation
doi.org/10.1016/j.energy.2018.05.163.
of different regions, Int. J. Hydrogen Energy 44 (56) (2019 Nov 12) 29390–29398.
[31] H. Scherrer, L. Vikhor, B. Lenoir, A. Dauscher, P. Poinas, Solar thermolectric
[4] S.R. Yan, R. Kalbasi, Q. Nguyen, A. Karimipour, Rheological behavior of hybrid
generator based on skutterudites, J. Power Sources 115 (1) (2003) 141–148.
MWCNTs-TiO2/EG nanofluid: a comprehensive modeling and experimental study,
[32] Y. Dou, F. Wu, L. Fang, G. Liu, C. Mao, K. Wan, M. Zhou, Enhanced performance of
J. Mol. Liq. 2 (2020 Apr), 113058.
dye-sensitized solar cell using Bi2Te3 nanotube/ZnO nanoparticle composite
[5] T. Zhu, E. Ertekin, Phonon transport on two-dimensional graphene/boron nitride
photoanode by the synergistic effect of photovoltaic and thermoelectric
superlattices, Phys. Rev. B 90 (2014), 195209, https://doi.org/10.1103/
conversion, J. Power Sources 307 (2016) 181–189.
PhysRevB.90.195209.
[33] M. Bharti, A. Singh, B.P. Singh, S.R. Dhakate, G. Saini, S. Bhattacharya, D.K. Aswal,
[6] T. Zhu, E. Ertekin, Phonons, localization, and thermal conductivity of diamond
Free-standing flexible multiwalled carbon nanotubes paper for wearable
nanothreads and amorphous graphene, Nano Lett. 16 (2016) 4763–4772, https://
thermoelectric power generator, J. Power Sources 449 (2020), 227493.
doi.org/10.1021/acs.nanolett.6b00557.
[34] E. Yin, Q. Li, Y. Xuan, Experimental optimization of operating conditions for
[7] K. Karthick, S. Suresh, H. Singh, G.C. Joy, R. Dhanuskodi, Theoretical and
concentrating photovoltaic-thermoelectric hybrid system, J. Power Sources 422
experimental evaluation of thermal interface materials and other influencing
(2019) 25–32.
parameters for thermoelectric generator system, Renew Energy 134 (2019) 25–43,
[35] S. Shittu, G. Li, X. Zhao, X. Ma, Y.G. Akhlaghi, E. Ayodele, Optimized high
https://doi.org/10.1016/j.renene.2018.10.109.
performance thermoelectric generator with combined segmented and
[8] M.M. Barry, K.A. Agbim, P. Rao, C.E. Clifford, B.V.K. Reddy, M.K. Chyu, Geometric
asymmetrical legs under pulsed heat input power, J Power Sources 428 (2019)
optimization of thermoelectric elements for maximum efficiency and power
53–66, https://doi.org/10.1016/j.jpowsour.2019.04.099.
output, Energy 112 (2016) 388–407, https://doi.org/10.1016/j.
[36] S. Fan, Y. Gao, Numerical analysis on the segmented annular thermoelectric
energy.2016.05.048.
generator for waste heat recovery, Energy 183 (2019) 35–47.
[9] U. Erturun, K. Mossi, Thermoelectric devices with rotated and coaxial leg
[37] S.R. Yan, M.Z. Mehrizi, L.K. Foong, Synthesis of NiAl-WC composite by the thermal
configurations: Numerical analysis of performance, Appl. Therm. Eng. 85 (2015)
explosion of elemental powders, Ceram. Int. (2020 Mar 9).
304–312, https://doi.org/10.1016/j.applthermaleng.2015.04.010.
[38] G.J. Snyder, Application of the compatibility factor to the design of segmented and
[10] Y.-X. Huang, X.-D. Wang, C.-H. Cheng, D.-T.-W. Lin, Geometry optimization of
cascaded thermoelectric generators, Appl Phys Lett 84 (2004) 2436–2438, https://
thermoelectric coolers using simplified conjugate-gradient method, Energy 59
doi.org/10.1063/1.1689396.
(2013) 689–697, https://doi.org/10.1016/j.energy.2013.06.069.
[39] S.C. Kaushik, S. Manikandan, R. Hans, Energy and exergy analysis of
[11] X. Wang, H. Wang, W. Su, F. Mehmood, J. Zhai, T. Wang, et al., Geometric
thermoelectric heat pump system, Int J Heat Mass Transf 86 (2015) 843–852,
structural design for lead tellurium thermoelectric power generation application,
https://doi.org/10.1016/j.ijheatmasstransfer.2015.03.069.
Renew Energy 141 (2019) 88–95, https://doi.org/10.1016/j.renene.2019.03.128.
[40] S.C. Kaushik, S. Manikandan, The influence of Thomson effect in the energy and
[12] M.A. Ahamat, M.J. Tierney, Timewise temperature control with heat metering
exergy efficiency of an annular thermoelectric generator, Energy Convers Manage
using a thermoelectric module, Appl Therm Eng 31 (2011) 1421–1426, https://
103 (2015) 200–207, https://doi.org/10.1016/j.enconman.2015.06.037.
doi.org/10.1016/j.applthermaleng.2011.01.002.

11
M.-W. Tian et al. Applied Thermal Engineering 181 (2020) 115996

[41] Bejan A, Tsatsaronis G, Moran M, Moran MJ, Moran MJ, Moran PGMCM. Thermal [47] M.W. Tian, A.G. Ebadi, K. Jermsittiparsert, M. Kadyrov, A. Ponomarev,
Design and Optimization. Wiley; 1996. N. Javanshir, S. Nojavan, Risk-based stochastic scheduling of energy hub system in
[42] A. Lazzaretto, G. Tsatsaronis, SPECO: A systematic and general methodology for the presence of heating network and thermal energy management, Appl. Therm.
calculating efficiencies and costs in thermal systems, Energy 31 (2006) 1257–1289, Eng. 1 (159) (2019 Aug), 113825.
https://doi.org/10.1016/j.energy.2005.03.011. [48] S.R. Yan, N. Shirani, M. Zarringhalam, D. Toghraie, Q. Nguyen, A. Karimipour,
[43] S. Asaadi, S. Khalilarya, S. Jafarmadar, A thermodynamic and exergoeconomic Prediction the boiling flow characteristics in rough and smooth microchannels
numerical study of two-stage annular thermoelectric generator, Appl. Therm. Eng. using of molecular dynamics simulation: investigation the effects of boundary wall
156 (2019) 371–381, https://doi.org/10.1016/j.applthermaleng.2019.04.058. temperatures, J. Mol. Liq. 19 (2020 Mar), 112937.
[44] H. Nami, A. Nemati, M. Yari, F. Ranjbar, A comprehensive thermodynamic and [49] H.S. Dizaji, S. Jafarmadar, S. Khalilarya, S. Pourhedayat, A comprehensive exergy
exergoeconomic comparison between single- and two-stage thermoelectric cooler analysis of a prototype Peltier air-cooler; experimental investigation, Renew.
and heater, Appl. Therm. Eng. 124 (2017) 756–766, https://doi.org/10.1016/j. Energy 131 (2019) 308–317.
applthermaleng.2017.06.100. [50] S.R. Yan, H. Moria, S. Asaadi, H.S. Dizaji, S. Khalilarya, K. Jermsittiparsert,
[45] F. Mohammadkhani, S. Khalilarya, I. Mirzaee, Exergy and exergoeconomic analysis Performance and profit analysis of thermoelectric power generators mounted on
and optimisation of diesel engine based Combined Heat and Power (CHP) system channels with different cross-sectional shapes, Appl. Therm. Eng. 13 (2020 May),
using genetic algorithm, Int. J. Exergy 12 (2013) 139–161, https://doi.org/ 115455.
10.1504/IJEX.2013.053387. [51] M.W. Tian, H. Moria, L.W. Mihardjo, A. Kaood, H.S. Dizaji, K. Jermsittiparsert,
[46] M. Antonik, B.T. O’Connor, S. Ferguson, Performance and design comparison of a Experimental thermal/economic/exergetic evaluation of hot/cold water
bulk thermoelectric cooler with a hybrid architecture, J. Therm. Sci. Eng. Appl. 8 production process by thermoelectricity, J. Cleaner Prod. 1 (2020 Jul), 122923.
(2016) 21013–21022.

12