Energy Conversion and Management 134 (2017) 260–277

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Review

A review on heat sink for thermo-electric power generation:

Classifications and parameters affecting performance
Ali Elghool a, Firdaus Basrawi a, Thamir Khalil Ibrahim a,⇑, Khairul Habib b, Hassan Ibrahim a,
Daing Mohamad Nafiz Daing Idris a
a
Energy Sustainability Focus Group (ESFG), Faculty of Mechanical Engineering, Universiti Malaysia Pahang, 26600 Pekan, Pahang, Malaysia
b
Universiti Teknologi Petronas, Department of Mechanical Engineering, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: In recent years, there have been growing interests in key areas related to global warming resulting from
Received 16 August 2016 environmental emissions, and the diminishing sources of fossil fuel. The increased interest has led to sig-
Received in revised form 6 December 2016 nificant research efforts towards finding novel technologies in clean energy production. Consequently,
Accepted 19 December 2016
the merits of a thermo-electric generator (TEG) have promised a revival of alternative means of producing
Available online 26 December 2016
green energy. It is, however, impractical to account for the cost of thermal energy input to the TEG which
is in the form of final waste heat. This is because the technology presents critical limitations in determin-
Keywords:
ing its cost efficiency nor its economic disadvantages. This paper reviews the principles of thermo-electric
Thermo-electric generator
Heat sinks
power production, as well the materials use, performance achieved, and application areas. The paper also
Waste heat takes a particular deliberation on TEG heat sinks geometries and categories. The review emphasizes more
Classification on the TEG performance while considering a number of heat sink parameters related to its performance.
Ó 2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
2. Recovery of waste heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
3. Thermo-electric properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
4. Thermo-electric material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
4.1. Semiconductor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
4.2. Ceramics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
4.3. Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
5. Thermo-electric generator (TEG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
6. TEG structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
7. Application of thermo-electric power generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
7.1. Low power generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
7.2. High power generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
7.2.1. Waste heat thermo-electric generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
7.2.2. Solar thermo-electric generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
8. Types of TEG heat exchanger (heat sink) and previous studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
8.1. Passive cooling heat sinks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
8.1.1. Metal plate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
8.1.2. Fin heat sink (Semi active heat sink) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
8.1.3. Fin to base attachment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
8.1.4. Analytical analysis of fin heat sink performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.1.5. Heat pipe (HP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
8.2. Active cooling heat sinks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

⇑ Corresponding author.
E-mail addresses: thamir@ump.edu.my, thamirmathcad@yahoo.com (T.K. Ibrahim).

http://dx.doi.org/10.1016/j.enconman.2016.12.046
0196-8904/Ó 2016 Elsevier Ltd. All rights reserved.
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 261

8.2.1. Fan-Fin heat sink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

8.2.2. Liquid cold plate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
8.2.3. Microchannel heat sinks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
9. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

1. Introduction high of 2000 [6]. Fig. 1 shows the published research on TEs
(thermo-electric generator and cooler) as a function of date from
With the rapid development of technologies and industrializa- Web of Science. The search subjects are thermo-electric (TE),
tion, energy demand will keep increasing. The modern economy thermo-electric and structure (TE + structure) and thermo-
relies largely on fossil fuels. This dependence is further fueled by electric and electron microscopy (TE + EM).
the advancement in the industrial sector and the adoption of However, low efficiency in conversion has remained to be a pri-
new technologies. However, fossil fuels have unique limitations mary challenge facing thermo-electric power production [7–9]. At
since their reserves are increasingly depleting. In addition, there present, researchers are making intensive effort to improve the effi-
are a number of environmental issues related to fossil fuels pro- ciency of thermo-electric materials. Table 1 shows studies that have
duction and use, at both local and global scales. Furthermore, a sig- been conducted on many types of thermo-electric materials, most
nificant portion of energy from fossil fuels is wasted during the of them during the last five years. At the same time, by the use of
transformation process and use. Owing to the issues related to fos- efficient heat sink, researchers can improve TEG performance since
sil fuels, energy experts around the world have expressed concerns effective dissipation of heat will keep temperature differences at
about the future of energy generation. Thermo-electric generator high levels. Thus, TEG performance can be maintained to operate
(TEG) has shown to offer a dependable and simpler way of at an optimum level. For this reason, this paper takes a review of
thermo-electric energy conversion. Other advantages of TEG mod- the classification of the heat sink and their application with TEG.
ule include lack of moving parts, environmental safety, and silent
operation. This generator can also be controlled in a seamless 2. Recovery of waste heat
and accurate manner. Researchers have shown increased interest
using thermo-electric technology in improving waste recovery effi- There are various reasons for developing more efficient pro-
ciency over the last three decades. This has been made possible cesses. These include the need to cut down on the budget, fossil
using a variety of heat-producing processes [1–5]. According to resources conservation, restricted legal framework, and scarcity of
the Web of Science database, annual publications in thermo- renewable resources. In the fields of energy research, political and
electric technology increased in the last 15 years from 500 to a economic actors, the topic of the significance of energy efficiency

25000

20000
TE
Researchers No

15000 TE+STRUCTURE

TE+EM
10000

5000

Years

Fig. 1. Number of published research on TEGs with search subjects being thermo-electric (TE), thermo-electric and structure (TE + structure) and thermo-electric and
electron microscopy (TE + EM).

Table 1
The conducted studies on many types of thermos-electric materials.

Isotropic Anisotropic With phase Pseudo-cubic Superionic With high band With Low lattice
layered transitions structures structures degeneracy thermal conductivity
PbTe [10–13] Bi/Sb2Te3 [14,15] GeTe [16,17] CuGaTe2 [18,19] Zn4 Sb3 [20,21] Half-Heusler [22–24] Skutterudites [25,26]
PbSe[27] In4 Se3 [28] SnSe [29] GuInTe2 [30] Cu2 Se [31,32] Stannites [33,34]
PbS[35,36] Ca3 Co4 O9 [37,38] Cu2 Se [31,32] Cu2S [39] Zintl phases [40,41]
SnTe[42] BiCuSeO [43] Cu2 S [39] AgBiSe2 [44–46] Clathrates [47,48]
Mg2 Si–Mg2Sn [49] SnSe[29] Ag2 Se [50]
SiGe[51,52] AgBiSe2 [44–46]
262 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

has become commonplace [53]. Researchers in these areas empha- two parameters are determined only by the electrical properties of
size the value of using waste heat in the production of space heating the conductor. Hence they usually combined to make the power
or electricity for the purpose of increasing value addition per each factor, PF (a2 r) [62]. On the other hand, the net thermal conductiv-
unit of energy used. By definition, waste heat includes sensible ity is a sum of two contributions: one from charge carriers (num-
and latent forms of heat which leak out of a system or which is bers of the holes and electrons transporting heat) and the other
not used by the system [54]. The sources of waste heat include from photons (value of photons that travel through the lattice).
waste water derived from washing, cooling or drying, refrigeration When a single type of charge carrier is major in a material, the total
systems, furnaces or automobiles as well as exhaust air emitted thermal conductivity ‘‘k” is the sum of the lattice thermal conduc-
from production halls [55]. The recovery of waste heat losses offers tivity ‘‘kl” and the charge-carrier thermal conductivity ‘‘ke”, that
an opportunity for low-cost and emission-free energy [56]. is, k = ke + kl [9,63,64]. The figure of merit can therefore be maxi-
Consequently, the industrial sector has continued to put more mized by increasing the electrical conductivity and reducing the
efforts in improving energy efficiency in its operation. About 25 thermal conductivity. A practical ZT of a suitable and high perfor-
percent of fuel energy is used in automobiles and accessories used mance thermo-electric material should be higher than 4. However,
in traditional combustion engines powered by gasoline. The to achieve such ZT remains to be a formidable challenge [65]. ZT is
remaining 75 percent is largely lost as waste heat [57]. It is clear important in determining the efficiencies of different thermo-
that the wasted energy will need to be harnessed and put to full electric generators working at similar temperature. A good TEG
utilization in order to save the energy and protect the environ- has high ZT, which is roughly 1, a result of improvement over the
ment. A novel device was developed to directly produce electrical years [66] .
power from heat. An example of such technologies is TEG which
can produce electricity while their power cycles are well estab-
4. Thermo-electric material
lished. Temperature originating from the source of waste heat is
dispensable to ensure efficiency in generating power. In general,
The TE materials can be classified into three catalogues: semi-
the generation of power from waste heat has remained limited
conductors, ceramics and polymers. Recently, certain polymers,
to only high and medium temperatures of the sources of waste
i.e. ethylenedioxythiophene, carbon fibre polymer-matrix struc-
heat. That is because technologies that create end use options are
tural composites, have also been shown to exhibit interesting
currently less developed and costly. Moreover, fluid and strong
thermo-electric material properties [67]. Deep descriptions of
components can condense as hot streams cool in the recovery
materials are not provided in this paper. Brilliant reviews of
equipment. This causes erosion and fouling conditions. The extra
thermo-electric materials, however, have provided explanation of
charge of materials that can withstand erosion and fouling regu-
both the material categorisations and the relationship between
larly prevents low temperature recovery uses [56].
material structure and thermo-electric properties [7,68–70].

3. Thermo-electric properties
4.1. Semiconductor
Thermo-electricity is the phenomenon of producing electricity
There has been an extensive push to enhance the figures of
from heat, and vice versa. The phenomenon is also referred to as
merit (ZT) of these materials to more than 3 to make them com-
‘‘Seebeck effect,” having been named after a German scientist, Tho-
mercially feasible. Semiconductor materials are favourable for
mas Seebeck. In 1821, Seebeck heated one end of two twisted
the development of thermocouples since they have Seebeck coeffi-
wires made of different metals. Thomas observed a small current
cients in excess of 100 lV K1 [71], and the best way to decrease
flow through the wires. Such flow is enhanced by the electrons
thermal conductivity without influencing electrical conductivity
which move from one end to another hence determining the direc-
and Seebeck coefficient in materials, thus enlarging ZT, is to utilize
tion of flow of the current. He concluded that electricity could be
semiconductors of high atomic weight. Alloys of Bi2Te3 with Sb, Sn,
generated from heat. Peltier discovered the reverse phenomenon
and Pb have high atomic weight. A high atomic weight decreases
later in 1834. Peltier found that when direct current went through
the speed of sound in the material and so reduces the thermal con-
the twisted wires, cooling and heating effect was observed from
ductivity [72]. Bismuth telluride is the most well-known type of TE
each side [58–61].
material. It is used with low and medium temperature. It is less
This paper seeks to explicate thermo-electric generator (TEG).
expensive than other thermo-electric material due to its mass pro-
Thermo-electric efficiency is a parameter that was derived by
duction and extensive variety of usage [9].
Altenkirch in 1911. It is called a thermo-electric figure of merit
Another communal type of semiconductor material is lead tel-
denoted simply as Z. Thermo-electric figure of merit is dimen-
luride PbTe [73]. It gains a thermo-electric figure of merit of 1.5
sional. To non-dimensionalize it, it is multiplied with the absolute
when covered with thallium at a temperature of 773 K. Together
temperature, T. The result of this is a dimensionless thermo-
with calcium manganese, lead thallium is used in thermo-electric
electric figure of merit, ZT, given as follows:
generators because they can handle high temperatures. This leads
ZT ¼ a2 rT=k ð1Þ to some TEGs being produced with segmented TE material. Lead tel-
luride is used on the hot side of the TEG because of its high ZT at
where a is the thermo-power or the Seebeck coefficient. The value a high temperatures while bismuth telluride is used on the cool side
is the voltage produced in every difference in temperature in a good of the TEG due to its high ZT at lower temperatures. Intermetallic
conducting material. The value r is the electrical conductivity of the compounds such as Mg2X (X = Si, Ge, Sn) and their solid solutions
conductor while k denotes the thermal conductivity. The three are semiconductors with the ‘‘ant” fluorite structure and have been
parameters closely depend on one another. Each makes up a part proposed as good candidates for high-performance thermo-electric
of the band structure. Particularly, a and r are inversely propor- materials. It has superior features such as high Seebeck coefficient,
tional to one another. This makes it hard to increase the thermo- low electrical resistivity, and low thermal conductivities [67].
electric figure of merit [9]. The two parameters, thermo power Together with calcium manganese, lead thallium is utilized as a
and electrical conductivity are mostly constant. A thermo-electric part of thermo-electric generators, in light of the fact that they
material that is of good quality has a low thermal conductivity, high can deal with high temperatures. This makes a few TEGs being pro-
thermo power and an increased conductivity of electricity. The last duced with segmented TE material.
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 263

4.2. Ceramics Table 2

Periods of development of TE materials [8].

Thermo-electric materials in practical applications are always Periods Conversion efficiency ZT

based on alloy materials, such as SiGe and Bi2Te3. Compared with Before 1950 <5% <1.0
TE alloys, metal oxides have advantages being less toxic and low 1950–1990 No exciting breakthrough –
cost, better chemical stability and oxidation resistance, so their After 1990 Up to 20% >2.0
use allows the production of more sturdy devices [74]. Ceramic
is a significant thermo-electric material for thermo-electric energy
thermoelectric phenomena in 1821, they are still far from the com-
generation to recover high-temperature waste heat from furnaces
petition. As a consequence, either more engineering of existing
or combustion engines [75]. Nevertheless oxides had not been
materials or discovering of completely new materials is crucial [8].
thought to be competitors as TE materials because of their low car-
The development of TEG conversion efficiency since TEG discovery
rier mobility, until the elite TE oxide of NaxCa2O4 showed up. Now
can be divided into three periods, as shown on Table 2.
cobalt-based oxides, such as Ca3Co4O9, NaCo2O4, have been fabri-
Nowadays, any successful thermo-electric development used
cated as p-type legs in TE modulus. As a counterpart, n-type SrTiO3,
domestically relies on the development of thermo-electric materials
ZnO and CaMnO3 ceramics have also been studied. Among them,
and thermal design of heat sinks. While the development of thermo-
CaMnO3 can be synthesized in atmospheric condition and shows
electric materials aims to increase efficiency and merit of the TEGs
brilliant TE properties, which makes the CaMnO3 a forthcoming
through novel materials, thermal design of the heat sinks aims at
competitor as an n-type oxide TE material.
increasing the transfer of heat by forming heat exchangers [91].

4.3. Polymers
6. TEG structure

Flexible thermo-electric apparatus requires a distinct class of

The TEG structure constitutes of a number of modules. Each mod-
polymeric materials known as ’’electronically conducting poly-
ule has an array of thermo-electric (TE) junctions that are connected
mers.‘‘ These are conjugated polymers with good electronic con-
electrically in series and thermally in parallel. These junctions are of
ductivity. Because of their special properties, electronically
p- and n-type TE materials. Accessories like cooling system, heat
conducting polymers are in high demand for the research and
absorber, electric insulator, electric conductor, and ceramic sub-
development of flexible gadgets in different areas. Samples of elec-
strate are auxiliary to the basic TE modules array. They add to the
tronically conducting polymers used include polypyrrole (PPY),
efficiency of the TEG. One module can generate 1 to 125 W of power.
PANI, polythiophene (PTH), poly (3, 4-ethylene dioxythiophene)
This can increase to 5 kW when modularly connected. The highest
(PEDOT), polyacetylene (PA), and their derivatives. These polymers
temperature difference (DT) between the hot and cold sides could
exhibit semiconductor features with an assortment of electronic
go as high as 70 °C [92]. In fact, the temperature difference can be
band structures. Among these, poly (2,7-carbazole), PEDOT, and
higher than 70 i.e., Lertsatitthanakorn et al. [93] have investigated
PEDOT: polystyrenesulfonate (PEDOT:PSS) have appeared as
the cook stove with its side wall adding a commercial TEG module
important candidates for thermo-electric applications. Their
made of bismuth-telluride based materials. The results revealed
advantages comprise good electronic conductivity, easy process-
that at a temperature difference of about 150 °C, TE module has pro-
ability, low thermal conductivity, low cost, formation of various
duced 2.4 W and the conversion efficiency is 3.2%.
sizes and shapes, and environment friendly [7,8].
The modules are solid in state and are devoid of movable parts.
The fact that they lack movable parts, there is a reduction in the
5. Thermo-electric generator (TEG) occurrence of wear and tear within the moving parts. This makes
them reliable and durable. They can have more than 100,000 h of
The increasing cost of energy, combined with global warming operation. The structure of the modules is simple. The operation
resulting from the use of fossil fuel have led to the growing efforts of the modules could be in two ways. They could be TEGs that pro-
to search for alternative energy sources that are both sustainable duce electrical power from a difference of temperature or thermo-
and clean [76–80]. TEG is one of the potential technologies used electric coolers (TECs) that create a temperature difference from a
in reducing cost and has drawn significant interest among direct current [94] which is not related to this review. The common
researchers. This is because TEG is a device which has the potential module structure is shown in Fig. 2 and 3.
to transform energy emitted from factories, power plants, comput-
ers, automobiles and human bodies into electricity by the use of
the Seebeck effect [1,81–85]. The merits of the Seebeck effect
include the absence of harmful wastes, operation in solid-state,
wide-ranging scalability, the extended life span of consistent oper-
ation, and its nature of maintenance-free operation because of
absence of chemical reactions or mechanical parts [86–89].
However, due to their nature of low efficiency, TEG materials
have attracted limited applications commercially. The only problem
that seems to be common with TEG is the inefficiency of the materi-
als, which makes it inadequate to complete with other traditional
methods of generating power [65]. In fact, the development of mate-
rial reaching a conversion efficiency to 20% happened in certain
applications of TE devices especially in the area of space missions
[8]. The NASA Jet had successfully conducted experiments in their
laboratories to develop advanced bulk materials capable of long-
term operation at temperatures up to 700 °C at more than 20% con-
version efficiency [90]. Although the performance of TE materials Fig. 2. (a) The device works as Seebeck effect (TEG), (b) the device works as Peltier
has been significantly encouraged since Seebeck discovered the effect (TEC).
264 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

Fig. 3. Practical TEG when P-N junctions connect in series to increase operating voltage.

7. Application of thermo-electric power generation mined by the dimensionless figure of merit of the TE materials
used and the generator’s operating temperatures, as shown below.
When classifying TEG applications for power generation, the pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
TEG can be divided into low power generation and high power gen- gte ¼ ððT h  T c Þ=T h Þ  ð 1 þ ZT  1Þ=ð 1 þ ZT þ T C =T h Þ ð2Þ
eration. Low power generation can generate power from 5 lW to where Th and Tc are temperature of the hot and cold sides,
1 W, where from 1 W and above the TEG is considered as high respectively.
power generation [9,87,95]. High power generation category con- The selection of the two temperatures is based on the use of the
sists of several TEG models connected together to generate big thermo-electric generators. Power generation involves the devel-
amount of power. Low power generation TEG generates a small opment of TEG waste-heat-recovery systems [97]. Such systems
amount of power to operate small devices such as watches, pace- have recently been developed by top automobile manufacturers
makers and hearing aids. such as BMW, Volvo, Ford and Volkswagen. This is aimed at
improving the fuel economy of these automobiles. Power is gener-
7.1. Low power generation ated from the thermo-electric generators in a range of 1 kW [40].
For a common gasoline-engine vehicle, around 40% of the fuel
The features of TEGs mentioned above make them developed as energy is released from the exhaust pipe and around 30% is lost
a stand-alone technology of power generation. These TEGs are used into the coolant. Effective utilization of these waste heat enhances
in biomedical, military, aerospace, and remote power applications. energy efficiency and saves money. In 1998, Nissan fabricated the
Electric devices that are incorporated in other bodies and use first thermo-electric power generator based on Si–Ge elements for
power-generation technology are grouped into mobile communi- automobiles [98]. The Bell Solid State Thermo-electrics (BSST)
cations category. They include iPods, MP3 players, and smart- group that includes BMW, Visteon, and Marlow Industries further
phones. Others are used in the medical field like cardiac made advancement in 2004 of an exceptionally efficient thermo-
pacemakers and hearing aids. The electric devices that are incorpo- electric system to recover waste energy from passenger vehicles
rated in other bodies have power requirements which range from [99]. Yang [100] showed that this system improved fuel economy
5 lW to 1 W. They have a life expectancy of up to 5 years [96]. by about 10%. In three years an owner of such car can save about
USD400 for a 8.32 km/l vehicle, under the supposition of
7.2. High power generation USD2.00 per gallon petrol cost and running distance of
24,000 km per year.
7.2.1. Waste heat thermo-electric generators Hsu et al. [5] had prepared a system for the recovery of waste
High power generation of TEG is mostly used in automobile heat that contains 24 TEG modules to generate electricity from
engines and industries. Iron and steel, chemicals, petroleum refin- the car exhaust pipe. This system was able to convert power output
ing, forest products, and food and beverage industries consume of 12.41 W at a temperature difference of 30 °C. In the industrial
enormous amount of energy, in which a large amount escapes to sector, the greater part of the late research on uses of thermo-
the environment in the form of exhaust heat. Table 3 shows electric power generation has been directed towards usage of
waste-heat source temperatures as a sample of mid- and high tem- industrial waste heat [87,101]. In this large-scale application,
perature TEG applications. The efficiency of TEG (get ) is deter- thermo-electric power generators offer a potential alternative of
generating electricity by industrial waste heat. This will contribute
towards solving the universal energy problem, and simultaneously
Table 3 help to some extent decreasing the global warming phenomenon.
Estimated waste-heat source temperatures of a sample of mid- and high temperature For example, in Thailand, Yodovard et al. [102] measured the
TEG applications [103].
potential of waste heat thermo-electric power generation for diesel
Application Source temperature Reference cycle and gas turbine cogeneration systems used in the industrial
range (°C) sector. Data was collected from 27,000 factories from different sec-
Automotive exhaust 400–700 [104] tors, namely; palm oil mills, petrochemical, oil refining, chemical
Diesel generator exhaus 500 [105] product, food processing, pulp and paper rice mills, textiles and
Primary aluminium Hall–Heroult cells 700–900 [106,107]
sugar. The thermo-electric power generation system was
Glass melting regenerative furnace 450 [106,107]
estimated to recover waste heat from the exhaust of cogeneration
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 265

systems by about 20% for the gas turbine, and 10% for the diesel 8. Types of TEG heat exchanger (heat sink) and previous studies
cycle, corresponding to a net power generation of about 100 MW.
A heat sink is a device whose work is dissipation or absorption
7.2.2. Solar thermo-electric generators of heat (thermal energy) from a hot surface to the gas, liquid or
These generators are known as STEGs (solar thermo-electric ambient air. With respect to the need for efficient heat dissipation,
generator). They were first designed to be used in aerial applica- heat sinks can be utilized in a variety of applications even to reduce
tions because of their long lives and reliability as well as their thermal deformation in sheet metal processing [112]. Some of the
capability of capturing high amount of solar radiation in space. common examples of heat sinks are cooling electronic devices,
They have recently been used widely in the harvesting of solar heat engine, and refrigeration. A metal device composed of series
energy in residential areas. This signifies the expanding utilization of fins known as a fin array is a common design of heat sink. In
of the STEGs through improving their popularity in other applica- order to increase the heat dissipation rate of the heat sink, it is nec-
tions [80,108]. This is because STEGs use a thermal collector incor- essary to either increase the thermal conductivity of fins, the heat
porated with the TEG which thus makes it more efficient. The transfer coefficient, or the surface area of fins [113].
thermal collector absorbs the solar heat from the sun. The heat is Different types of heat sinks can be used by a thermo-electric
focused to one point and then conducted through the TEG using generator. This is with respect to reducing the cost of materials,
a fluid pipe or use of other method. This makes the thermal resis- ease of manufacturing and installation, and enhanced heat flux
tance of the TEG cause a temperature gradient that is directly pro- density. There is a generation of high open circuit voltage by TEG
portional to the heat flux generated from the thermal collector’s when there is a large temperature gradient between the hot and
absorber to the fluid. The consequence of this is a generation of cold sides. If a large heat flux is applied on the TEG hot side and
electric power by the TEG that is proportional to the temperature the cold part is maintained at low temperature then high conver-
gradient [80]. So far, the efficiency of sunlight based TEG systems sion efficiency can be achieved. An effective heat sink is required
achieved is around 5% for a temperature difference of about in this case. The efficiency of TEG also highly depends on good
100 °C and ZT materials of 1 [109]. Therefore, the main commercial selection and careful designing of the heat sink. Table 3 indicates
obstacle of solar TEG is its 5% conversion efficiency, making it the general classification adopted by TEG heat sinks.
somewhat avoided as compared to other types of solar electrical Currently, there are five categories of the cooling procedure of
technologies such as solar cells which exhibit an efficiency of thermo-electric heat sink, namely; phase change cooling, natural
18%. Higher conversion efficiencies are expected in the future. This air cooling, forced air cooling, edge cooling and liquid cooling
is linked to the development of improved ZT of the thermo-electric [88]. These procedures can also be grouped into active cooling
materials used, and improved development in the design of solar (phase change cooling and forced air/ water cooling) and passive
collectors [70]. cooling (natural air convection and heat pipe) [45]. This section
Milijkovic et al. [110] have studied a solar thermo-electric of the paper intends to focus on the grouping of these cooling
(HSTE) system using passive heat transfer to the bottom of the sys- procedures including the studies conducted, together with a sum-
tem by the use of heat pipe. The mirrors focus solar energy onto the mary of the parameters utilized in these studies as shown in
hot side of the TEG. As indicated by the researchers, the develop- Table 4.
ment of the HSTE can achieve a conversion efficiency by as high
as 52.6% when the sun concentration is focused to an equivalent 8.1. Passive cooling heat sinks
of 100 suns with temperatures of the bottoming cycle at 413 °C.
Zhang et al.[111] investigated the solar based thermo-electric 8.1.1. Metal plate
cogeneration (STECG). The outcome demonstrates that for a figure This is a flat metal piece that is usually made from copper or
of merit of 0.59 and sun based insolation of below 1000 W/m2, the aluminium. It is usually attached on the cold side of the TEG for
system can produce 0.19 kWh of electrical power and around 300 L heat dissipation through natural convection where it is not depen-
of warm water at 55 °C in one day and the collector efficiency, dent on air flow as indicated in Fig. 4. This type of heat sink has a
power output and electrical efficiency can increase up to 47.54%, limited capability with respect to power dissipation and rare when
64.80 W and 1.59%, respectively. utilized with TEG at applications with low power density.

Table 4
Classification of heat sink groups [114].

Heat sink Advantages Disadvantages Applications Examples

category
Passive User friendly Power dissipation capacity is Natural convection/ Metal plate
Readily available limited systems
Cheap Not dependent on air
flow
Simple power intensity
uses
Semi-active In relation to first one it has low thermal resistance Limited in capability of power Medium power Fin heat sink
for the same area dissipation intensity uses
Active Combine fan and heat sink into one unit Long term reliability High power intensity Fan-fins heat sink
Produce much heat dissipation capacity in relation to Cost uses
previous types
Liquid cooling Produce much heat dissipation capacity more than Complicated High power intensity Liquid cold plate
all previous types Expensive uses
Need constant heat
circulation
Phase change Heat is distributed equally Complicated High power intensity Vapour compression phase
cooling The cooling fluid fast dissipate heat Expensive uses change cooler
Need extra area
266 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

(a) Natural convection (b) Forced convection

Fig. 4. Heat convection from a plate surface.

8.1.1.1. Mathematical analysis of plate heat sink. To model the con- tion to the other two. The heat flux resulted in 18,125 W/m2 for A
vection heat transfer on the flat metal piece we can use Newton’s and 31,195 W/m2 for B when the speed of ambient wind was 0 to
Law of Cooling and apply it on Fig. 4. 5 m/s.

Q ¼ hAðT surface  T free Þ ð3Þ 8.1.2. Fin heat sink (Semi active heat sink)
This is a fin array made from metal. There are round, triangular
where and rectangular pin fins in the profiles of the longitudinal fins. The
rectangular one is mostly used among these profiles alongside
Q = Heat multiple fin arrays [113]. The round pin heat sinks are character-
h = Convection coefficient ized by omni-directional features and are mostly preferred with
A = Surface area low airflow or when its direction of flow is unknown. Both types
Tsurface = Temperature at surface are illustrated in Fig. 5. TEGs are usually known to be composed
Tfree = Temperature in free stream. of used air heat exchanger straight fins which translate into most
utilized heat dissipation system in thermo-electric influenced by
The typical length and Rayleigh number for the natural convec- its low cost of manufacturing [91]. The extruded heat sinks are
tion can be represented in the following equations: usually the cheapest fins available from many companies.
Usually, the extrusion process fails to produce thin and closely
As
Lc ¼ ð4Þ spaced fins. This prompts some companies to manufacture their
p heat sinks with the help of grooves that machine a flat plate and
an epoxy resin is used to stick the fins to the plate as shown in
gbðT s  T 1 ÞL3c Fig. 6. There is the capacity to obtain a higher fin surface although
RaL ¼ Pr ð5Þ
v its negative effect is the thermal resistance of the heat sink base.
Some manufacturers do give information on the performance of
the fins. Straight fins as manufactured by shaving process are
Nu ¼ 0:54RaL1=4 ð6Þ referred to as ‘skyving’ [116]. Its limited capability of power dissi-
A comparison has been made between the three methods used pation often makes it non-considered for application with TEG.
for the cooling of TEG’s, namely; metal plate heat sink; 100 mm fin,
80 mm fin and 60 mm finned heat sink; and heat pipe with a 8.1.3. Fin to base attachment
finned condenser at a speed of wind ranging from 0 m/s to 5 m/s It is important that the processor material utilized in the attach-
[115]. Two forms of TEG were used with distinct favourable hot ment of the fins has low thermal resistance. The best solution at
side temperatures; A at 150 °C, while B at 250 °C. The metal plate this point is the material possessing continuity. It is expensive to
cooling heat sink had the capability to reach lower heat flux in rela- machine fins from a thick plate. Aluminium or copper is preferred

(a) Rectangular heat sink (b) Round pin heat sink

Fig. 5. Diagrams illustrating rectangular and round pin heat sink.
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 267

(a) Extruded fins (b) Stacked fins

Fig. 6. Diagrams illustrating the extruded and stuck fin.

in the making of the fins due to their lower weight. In addition, The overall effectiveness is given by:
copper has the ease of bonding as compared to aluminium [116].
Q total;fin Q u þ Q fin
A thermal interface material (TIM) is utilized between the heat efin;ov erall ¼ ¼ ð13Þ
Q total;nofin hAb ðT b  T f Þ
sink and component surface to increase the heat flux from the hot
surface to the heat sink. Thermal grease and thermal pads are both The fin efficiency is:
common interface materials that come with the most appropriate
Qf 1
thermal performance. This is because a thin surface is created gfin ¼ ¼ ð14Þ
between the heat sink and the package after grease occupies all hAfin;max ðT b  T f Þ mL
the air gaps. This results in low thermal resistance with the range The temperature is increased to T2. The heat, Q, transferred from
from 0.1 to 0.2 °C/W. Graphite foil has also been recently applied the heat sink [19] is:
for use with TEG. The material is easy to apply and has excellent
performance. It has the capacity to be used at high temperatures. Q ¼ hAðT s  T a Þ ð15Þ
Fig. 7 illustrates a microscopic Look at Surfaces (a) without TIM The heat transfer from the heat sink absorbed by air is given by:
and (b) with TIM.
_ p ðT 2  T 1 Þ
Q air ¼ mC ð16Þ
8.1.4. Analytical analysis of fin heat sink performance The thermal resistance of the fin can be calculated as:
In this case, the maximum heat transfer from the fin can be
expressed as [16]: 1
Rfin ¼ ð17Þ
gf Afin hfin
Q finmax ¼ Afin hðT b  T f Þ ð7Þ
The thermal resistance of the base material is given by:
Total heat fluxed is that from the un-finned surface plus the
heat flux from the [17]: 1
Rbase ¼ ð18Þ
Abase hbase
Q T ¼ Q u þ Q fin ð8Þ
The combined thermal resistance of fin and base material can
The heat transfer rate without the fin from area A to the sur-
be expressed by:
rounding fluid is [18]:
1
Q u ¼ Au hðT b  T f Þ ð9Þ Rf ¼ ð19Þ
1
Rfin
þ R1
base
The heat transfer rate with very long-fin is:
where Rf explains the conductive part of the thermal resistance in a
ðT b  T f Þ tanhðmLÞ ¼ gfin  Q fin;max
1=2
Q fin ¼ ðhpkAcÞ ð10Þ heat sink.
The convective contribution can be simplified as:
 1=2
hP 1
Since m ¼ ð11Þ Rflow ¼ ð20Þ
kAc _ p
mC
The fin effectiveness can be determined by this relation: One could consider the spreading resistance as well by the
Q fin definition:
efin ¼ ð12Þ The contribution of the spreading resistance to the overall
hAb ðT b  T f Þ
device temperature rise is significant if the footprint of a heat sink

Fig. 7. A microscopic Look at Surfaces (a) without TIM and (b) with TIM.
268 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

is larger than the size of the heat source. Considering this, the total section from the evaporator. Due to the condenser section of the
thermal resistance of a heat sink can be defined as: HP being in a cooler ambient, its surface is slightly cooler. Vapour
contacting with the surface of condenser then condenses and it
Rt ¼ Rf þ Rflow þ Rspd . . . : ð21Þ
moves heat through vapourization which is then dissipated to
The transfer of heat in the fin heat sink was modelled by analyt- the environment. The liquid goes back to the evaporator section
ical method while a mathematical model with a finite element of the HP via the wick due to capillary force which then completes
software was utilized for the prediction of the TEG performance the cycle. There is the absorption of heat at the end of the HP and
[117]. The initial optimization step is where the optimum fin spac- rejection takes place on the other side while the fluid inside acts as
ing belonging to a certain heat sink geometry tallies with respect to a medium for the transport of heat. The condensation and boiling
the analytical approach. The second optimization stage is also processes are related with high coefficients of heat transfer. It is
referred to as a compromise programming and utilizes the design therefore usual that the HP is an effective device of heat transfer
approach in which the length of the heat sink is reduced by enlarg- due to alternate condensation and boiling of the working fluid
ing the fin tip. Based on a found compromise point through the being the basis for this operation.
heat sink efficiency, there is a reduction by 20.9% in relation to The thermal conductivity of heat pipe depends on the design of
when there is no optimum design and the output power of TEG the heat pipe itself. In a heat pipe where water is used as the work-
is improved by 88.7%. This is why it is appropriate to be applied ing fluid there is an effective thermal conductivity of 100,000 W/
as a heat sink. Scaling down the TEG can additionally improve m K as compared to 400 W/m K for copper. Heat pipe usually
the TEG power density when the length of the heat sink is below reaches an effective conductivity of 400,000 W/m K, a thousand
14.5 mm. times more in relation to copper. A heat pipe with water that is
0.6 cm horizontal diameter and 15 cm long has the capacity of
8.1.5. Heat pipe (HP) transferring heat at a rate of 300 W [113]. The lightweight simplic-
This is a device that does not have moveable parts with the ity and construction in addition to a variety of materials, shapes
capacity of transferring high quantity of heat over relatively long and sizes enables a fine tuned performance of the HP to be applied
distances significantly at a fixed temperature without any power in a wide range of temperatures and applications [119,120]. During
input. The best TEG heat exchanger with respect to medium- HP designs, it is essential to consider the compatibility of materials
range temperature under 300 °C is heat pipe heat sink (HPHS), due to the presence of multiple interfaces (working fluid, wick,
according to Energy Efficiency Guide for Industry in Asia [118]. wall and other fluids in contact with the HP).
The heat pipe (HP) is a closed pipe composed of a wick placed on Although HP designs can be done by a variety of materials, the
the inside of its surface and little quantity of liquid like water pre- following materials are commonly applied for components:
sent at its saturated state as indicated in Fig. 8. It has three parts;
firstly the evaporator part on one end where absorption of heat 1. HP wall (or shell): titanium, aluminium, nickel stainless steel
takes place in addition to vapourization of the fluid, secondly the and copper.
end of the condenser section where condensation of vapour and 2. Working fluid: silver, sodium, ammonia, methanol, hydrogen,
rejection of heat takes place, and thirdly the adiabatic part in the helium and water.
middle where the liquid and the vapour stages of the fluid stream 3. Wick: cloths, carbon fibres, titanium, bronze, nickel, stainless
in reverse ways starting from the core going through the wick. The steel and copper [119,120].
cycle is completed without substantial transfer of heat between
the surrounding medium and the fluid. Thermodynamic properties It is essential to use appropriate combination of fluids and
are the basis for HP’s operations in fluids undergoing vapourization materials within the HP in addition to the ones in external contact
at one end and condensation at the other. It starts with the satura- with the HP. This facilitates for durable designs with a long
tion of the HP’s wick with fluid and filling of the core part with lifetime.
vapour as indicated in Fig. 8. Contact between a hot surface or Choosing the appropriate wick configuration is dependent on
hot environment with the evaporator end of HP causes heat to flow the application and is based on the cost, performance and compat-
into the pipe. Since most of the liquid at the end of the evaporator ibility of material. The following are typical wick configurations;
at a saturated state, heat transfer will cause vapourization result- foams, felt (fibrous media), sintered powders, grooves and screens
ing more vapour pressure. The resulting differences in pressure [119,120,122]. High heat flux is facilitated by the sintered wick in
force the vapour via the core of the heat pipe flow to the condenser addition to the wide angle of work, hence, being recommended for

Fig. 8. A cross section of a heat pipe.

 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 269

Fig. 11. Sample heat pipe model [124].

For a heat sink with nf fins, fin spacing of s, fin thickness of t,

and fin depth of p, the heat will be:
2p þ t
Fig. 9. Common wick structure: Sintered, Mesh (screen) and Grooved [121]. Q fin ¼ ð24Þ
2nfP þ L

a variety of applications. The three most common types are illus- S

Q base ¼ ð25Þ
trated in Fig. 9. Types of HPs include; inverted meniscus, micro 2nf þ L
and miniature HPs, pulsating HPs, capillary-pumped loop HPs, loop
On the surface of a heat sink, heat is transferred to the sur-
HPs, gas-loaded HPs, rotating HPs, vapour chamber HPs (flat
rounding by two parallelmodes: (1) natural convection and (2)
shaped), annular HPs, capillary-driven HPs (sealed pipe composed
radiation [125]. The surface thermal resistance of a heat sink can
of a wick HP) and two-phase closed thermo-syphon (sealed pipe
be defined as:
without wick TSs) [119,120]. Rectangular or cylindrical are the
most common cross-sectional geometries for HPs. However, there DT
R¼ ð26Þ
are different designs and shapes that work with respect to specific Q convection þ Q radiation
purposes. There is fabrication of the micro HPs in cross-sectional
where DT is the temperature difference between the surface of the
dimensions estimated at 10 lm [119].
heat sink and the surrounding. The amount of convective heat
On the other side, there is also fabrication of the conventional
transfer is the sum of heat transfers from the fins and the base.
large scale HPs with length in the order of 100 m [120]. The selec-
tion of material dictates the range of operational temperature and Q convection ¼ hf Af DT þ hb Ab DT ð27Þ
shifts from -200 °C to very high temperatures of up to 2000 °C. TSs/
HPs are highly reliable and durable lasting for more than 13 years where f refers to ‘‘fin” and b refers to ‘‘base plate”. The following
without showing signs of deterioration [123]. relations are used for calculating the convective heat transfer coef-
ficient in rectangular finned heat sinks:
8.1.5.1. Modelling heat pipe heat sink. In a heat pipe, heat transfer k
occurs from the heat pipe heat sink to the surrounding by three hb ¼ 0:59Ra0:25
L base plate ðsingleÞ ð28Þ
L
paths: (i) inside hollow, through conduction and convection, (ii)
" #0:5
wick, through conduction, and (iii) wall, through conduction.
k 576 2:873
hf ¼ þ parallel plates ðfinsÞ ð29Þ
s Ras s2 Ras s0:5
Fig. 10 illustrates the thermal resistance circuit of a heat pipe.
The radial thermal resistance for a cylinder wall with inside and L L
outer radii of r1 and r2 and length of L with thermal conductivity of
k is: gb
Rax ¼ ðT s  T amb Þx3 ð30Þ
r2 ma
Ln r1
Rr ¼ ð22Þ !
2pkl   X 1
Q Radiation ¼ r T 4s  T 4amb 1ei
ð31Þ
Thus, the resistance of a cylindrical heat pipe is simply summa- i Ai
þ 1
Ai F i1
tion of all the resistances shown in Fig. 11.
0    1 Remeli [118] used a new way of recovering waste heat to elec-
r out
Lc þ Le @ln rwall ln rrwall tricity using heat pipe heat sinks and TEG. A numerical model was
RHP ¼ þ
wick
A ð23Þ
2pLe Lc kwall kwick established to predict heat transfer rate demonstrating how effec-
tive the heat exchanger system was. There was an experiment that
utilized a laboratory scale system for the validation of the theoret-
ical model. The experimental results and the prediction were in
accordance since the same trend was present in both results. With
respect to the experimental results, there was an increase in the
effectiveness of the heat pipe from 67.9% to 72.4% with the increase
of air face velocity. Remeli [126] conducted another experiment
based on assessment of the impact of ratio of mass flow rate on
the heat transfer. The outcomes indicated that a higher mass flow
rate in the upper duct to lower duct ratio exhibits a favourable
impact in the entire performance of the system. A higher output
and higher amount of heat transfer are influenced by a higher mass
flow rate ratio. Air face velocity and the rate of heat transfer are the
parameters utilized by the authors in the evaluation of the effec-
tiveness of the HPHS system. It was evident that a rise in air face
Fig. 10. Diagram illustrating heat pipe thermal resistance [124]. speed was proportional to an increase in the effectiveness of HPHS.
270 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

Orr [127] conducted an experiment in addition to the calculation of efficiency of the system (unorganized dissipation of heat) and
the efficiency of the system on thermo-electric cells with the help noisy operations [130]. Considering the inherent low conversion
of an engine exhaust and heat pipes. The eight cells utilized had the efficiency of thermo-electric generators, heat sinks that lack any
capacity to generate 6.03 W for battery charging. The efficiency of form of auxiliary power consumption are mostly preferred.
conversion of heat to electricity throughout the system was 1.4%.
The difference between the expected efficiency and the actual effi- 8.2.1.1. Theoretical analysis of fin heat sink. The thermal resistance
ciency was 2.3% and this might have been influenced by the cells of the finned heat sink and the heat flux applied across the heat
failing to work at maximum voltage. The expected efficiency sink is given as:
equals roughly 11.0% of the Carnot efficiency while the real effi-
1
ciency is 6.7% of Carnot efficiency. Rhs ¼ ð32Þ
hhs  ðAb þ Nfin :gfin :Afin Þ
An experimental and theoretical analysis was utilized while
determining the heat dissipated for passive heat sink with TEG The fin arrangement and the convection heat transfer coeffi-
[115]. This research involved investigations of two commercially cient represent the most effective parameters on the thermal resis-
available TEGs (Type A and Type B) with hot side temperatures tance. The selection of the optimum number of fins and fins gap of
(150 °C for Type A, and 250 °C for Type B). The maximum experi- the heat sink for certain air speed along the length of the fins help
mental and theoretical heat flux was determined for both TEGs. to design optimum performance of the fins. Fin gap is illustrated in
The following are the conventional approaches utilized in the cool- Eq. (33).
ing of TEGs: heat pipe with finned condenser, finned block and
w  Nfin :t fin
bare plate. Air speed at the TEG’s cold side ranges from 0 to 5 m/ C¼ ð33Þ
s. The heat pipe finned condenser achieved a maximum limiting Nfin  1
heat flux of 40,375 W/m2 for Type A, and 76,781 for Type B. where w is the width of the heat sink and t fin is the fin thickness as
Ashwin Date et al. [125] proposed a system with the capacity to shown in Fig. 13. The exposed surface area Ab of the base plate is
generate hot water and electricity from TEG. Solar was utilized as defined in Eq. (34):
the source of heat while the heat pipes were utilized in the dissipa-
tion of heat from the TEG. The dissipated heat was then utilized for Ab ¼ ðNfin  1Þcdotc  l ð34Þ
domestic heating of water. A governing equation for this proposed Total surface area Afin of one fin including both sides is defined
system was developed by a theoretical analysis. An experiment in Eq. (35):
was then set up in the laboratory to verify the theoretical analysis.
With heat source of 50,000 W/m2 the thermo-electric generator Afin ¼ 2  xfin  l ð35Þ
achieved a temperature difference of 75 °C to 80 °C of hot water Velocity ðv Þ of air flowing over the fins can be calculated from
usable in industrial and domestic applications. equation 3.5 by knowing the volume flow rate ðv_ Þ:
v_
8.2. Active cooling heat sinks v¼ ð36Þ
ðNfin  1Þ  c  xfin
8.2.1. Fan-Fin heat sink Teertstra’s equation used to estimate the Nu for determination
All the elements mentioned on the previous fin heat sink type of convection heat transfer coefficient (h) is being used here and
apply to this fan-fin type too with exception to the use of its fan. mentioned below [131]:
Heat sink and forced convection are integrated in one unit. With 2 30:33
relation to the semi-active and the passive heat sinks it has a
6 1 1 7
higher capability to produce higher heat dissipation. Fig. 12 shows Nu ¼ 4 þh pffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffii3 5 ð37Þ
RePr 3
the pictures of a fan-fin heat sink. The air inflow through the heat 2 0:644 Re  Pr0:33 1 þ 3:65 pffiffiffiffi
Re
sinks can be categorized into bypass and non-bypass with respect
to their characteristics. The bypass flow involves air flowing
through the top of the heat sinks while the air passes through
the channels between the fins in the non-bypass flow [128]. With
respect to the low specific heat capacity of air, the fan cooling
approach does not have the capacity to achieve high temperature
difference [5].
Other shortcomings of this fan-fin approach include: low relia-
bility (due to short fan life), extra consumption of electricity, low

Fig. 12. Pictures for fan-fin heat sink [129]. Fig. 13. Schematic of dimensions of heat sink [131].
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 271

The fin efficiency can be calculated from the following formula: cial module in addition to the comparison of the performance of
the new and commercial modules. There is also the highlighting
tanhðm  xfin Þ
gfin ¼ ð38Þ of the impact of the parameters of thermo-electric on electrical
m  xfin power output. The commercial module for this application
While m is as defined in Eq. (39) as: required 8 fins, 100 W/m2 K of heat transfer coefficient and
40 mm of fin height as the appropriate parameters. The materials
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2h best suited for temperature ranges of 300–573 K with respect to
m¼ ð39Þ the analysis were TAGS, lead telluride and bismuth telluride from
K fin  tfin
three distinct thermo-electric groups namely; n-type, hot forged
Gou, Xiao studied the effect of heat transfer irreversibility on Bi2Te3 and p-type (Bi, Sb). Comparisons were made with the output
TEG performance [132]. Comparison of the rate of conduction electrical power of other models as well as the commercial mod-
was made between the theoretical system model and thermo- ules for the thermo-electric materials that have been recently
electric generator setup with respect to the basic effects of TE developed. The commercial module produces a maximum electri-
and the irreversibility of heat transfer. According to the results, cal power of 3.7 W, however, more recently developed TE materials
there are other applications other than the increase of waste heat achieved a maximum electrical power of 4.4 W.
temperature and TE modules in series, influencing the capacity of Tzeng and Jeng conducted a series of systematic and successful
heat transfer on the cold side in an appropriate range and maxi- experiments [135] that were based on determining the electricity
mizing the surface area of the heat sink in the appropriate range. generated by a TEG system from exhaust pipe under different
It can also be used in boosting up the performance of this setup. operating conditions such as the flow rate of cooling air
The performance of the thermo-electric generator is affected by (FC = 0–336 m3/h), the flow rate of hot air (FH = 3–15 m3/h), the
irreversibility of heat transfer while irreversibility in the transfer inlet temperature of hot air (Ti = 100–350 °C) and the configura-
of heat on the cold side delays the performance of the system. tion of the heat absorber. There was a proposal of a one-
With respect to Astrain, Vian [133] there was a computational dimensional steady heat conduction model that would have the
study based on the effect of the thermal resistance of heat sink Seebeck effect and an internal Joule heat generation mechanism.
(in both cold side and hot side) on the power output of a TEG where The results of the experiment verified that the proposed theoret-
it involved the development of a computational model. The model ical model was valid. Critical heat transfer parameters on electric-
utilizes a numerical approach in the simulation of the performance ity generation with respect to integral thermo-electric generator
of the TEG where both heat source and heat sink are included. A system must be analyzed as reference for any new design of
prototype was constructed and tested in order to verify the preci- physical product.
sion of the computational model. A simulation was made on the
application of smoke heat from the combustion chimney for TEG. 8.2.2. Liquid cold plate
The outcomes showed need for a meter of chimney height to pro- The most efficient approach for removal of heat is liquid blocks
duce an estimated 1 kW of electric energy translating into 280 W/ (cold plates). They are often 4–5 times more efficient compared to
m2. With respect to this study, thermal designs are useful in these a standard air-cooled heat sink and fan. Usually (provided the liquid
forms of applications. The electric power generation is increased is ethylene glycol/water or typically water) the pump liquid is used
by 8% with an increase in thermal resistance in both heat exchang- again in the liquid block. The liquid is moving back and forth through
ers by 10%. It was also confirmed that the same advancement in some of the water block chambers which allow movement of heat
both thermal resistances should be achieved to increase the gener- from the cool plate of TECs. As Fig. 14 indicates, the warmer liquid
ation of electric power. A heat sink must be designed to enhance the is then removed from the water block. The liquid source can be:
performance of the thermo-electric generator.
There was a study by M.C Barma et al. [134] on the impact of i. Building (tap) water – It is cooler in relation to ambient air.
the parameters of a plate fin heat sink on the differences in tem- The removed warm water can be used in other ways or sent
perature, TEG efficiency and output electric power for the commer- to the drain.

Fig. 14. Schematic of multiple branch channels heat sink [130].

272 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

ii. Chiller liquid – A pre-chilled liquid (dielectric fluid like fluo- fluid surface for a long flat path was much larger than that for
rinert, water/glycol, oil or typically alcohol) passing through shorter paths split into two fluids.
the water block from the recirculating chiller where it then In future, it is essential to consider the pumping power during
returns for re-cooling. This source is usually a closed loop the design of the TE generation system. A study was carried out
system. by Niu and Yu [141] for the development of a TEG unit utilizing
iii. Stream water, lake or well water – There is need for pre- commercially available thermo-electric modules that are based
filtering in this source; however, it is significantly cooler on Bi2Te3 for generation of a maximum power of about 150–
than ambient air temperature. The removed warm water 200 W level. The objective of the study was the validation of the
can be reused or returned to the source. numerical model that was previously published in addition to pro-
iv. Radiator closed loop – A liquid (propylene glycol, water, or a vision of some guidelines for modifications of the model. There is
mixture of both) goes into the water block at temperatures construction of a thermo-electric generator unit with parallel plate
estimated to be similar to ambient air. The warm water com- heat exchanger fabricated purposely for the experiments. The tests
ing out goes through a fan cooler radiator of substantial size are intended to assess the impact of the major working circum-
that cools it and returns to the water block representing a stances including the temperatures of the cold and hot fluid inlet,
closed loop system. the rate of flow and power output resistance and TEG efficiency.
The results verified that both conversion efficiency and maximum
There are additional auxiliaries required by this type such as
power output are substantially influenced by the circumstances of
water treatment, valves, pump and pipes, hence being perceived
operations using specifically the temperature of the hot fluid inlet
as sophisticated and expensive. Thermo-electric generators have
and the rate of flow. The preliminary validation of the numerical
low conversion efficiency and hence are not advisable to utilize liq-
model for comparison, the level of accordance between the mea-
uid cold plate for thermo-electric generators that may need auxil-
sured data and numerical model is encouraging, hence showing a
iary power consumption.
direction for advancement of the numerical model.

8.2.2.1. Analytical analysis of water block heat sink performance. By

8.2.3. Microchannel heat sinks
considering the half section of the cold plate the resistances can
The idea of a smaller scale channel was initially proposed by
be calculated. First, the heat is transferred from the Traveling-
Tuckerman and Pease [104] in heat sinks around three decades
Wave Tube (TWT) to the cold plate through conduction and con-
back. They expected that an increase in heat transfer coefficient
ductive resistance is calculated as [136].
could be done with the decrease in channel hydraulic diameter.
t This improvement gave a flash in the electronic engineering which
Rcond ¼ ð40Þ
kAb was then handling issues of high heat dissipation in restricted
space. Micro- and mini-channels are not the same as the conven-
Spreading resistance is calculated by the relation as follows tional channels in terms of channel hydraulic diameters [142].
[137]: Micro-channels can dissipate high heat densities, even more than
1000 W/cm2 [143] unlike the traditional channels which can
Wmax
Rsp ¼ pffiffiffiffi ð41Þ remove up to 20 W/cm2 of heat flux [144]. Thus, discovery of this
kr1 p innovation was exceptionally valuable. Arrangements in terms of
channel hydraulic diameter suggested by Mehendale et al. [145]
where Wmax is the constant and r1 is the equivalent radius of rectan-
and Kandlikar and Grande [146] are generally used, the latter being
gular source.
more common. The classifications assumed by these authors are
Nusselt Number (Nu) is given by Dittus-Boelter equation [138]:
displayed in Table 5.
Nu ¼ 0:023Re0:8  Pr0:4 ð42Þ A cover is used to insulate the top of the heat sink which is con-
sidered as adiabatic. The micro channels of the heat sink are
The convective resistance which is a prominent component in
pumped with liquids such as water to allow for extraction of heat
the total resistance network analogy is given by:
on the hot surface. Tuckerman in the 1980 s viewed laminar flow
1 as favourable for the removal of heat through a micro channel
Rconv ¼ ð43Þ
hAs due to the thin thermal boundary layer that developed. Micro
The total thermal resistance (Rt) from the TWT surface to the channel heat sinks were designed with distinct dimensions to
fluid temperature is expressed as follows [139]: study the heat transfer and flow friction within these sinks. An
optimization procedure was also developed for the prediction of
Rt ¼ Rcond þ Rsp þ Rconv ð44Þ
the best aspect ratio of the channel to reach the most favourable
A numerical calculation was conducted in the evaluation of the heat transfer. Water was utilized as the working fluid in the tests.
thermo-electric power generation with respect to heat transfer and The head loss was noticed to be more than usual. This was in addi-
the fluid dynamics [140]. FLUENT, a commercial software under- tion to the thermal resistance Dt/qw with respect to a single
went coupling with the TE model was used purposely for this dimensional heat flow being 20 times lower in relation to other
objective. Evaluation of the electromotive force could be made heat sink utilized for cooling in combined circuits (see Table 6).
for split joining flow and counter flow model. An impact of stagna- Since the work of Tuckerman in 1980s, there had been much
tion point was discovered at the latter model. The friction along the investigation with the intent of further comprehending the

Table 5
Mehendale et al., Kandlikar and Grande channel organization.

Mehendale et al. Kandlikar and Grande

Conventional channels Dh > 6 mm Conventional channels Dh > 3 mm
Compact Passages 1 mm < Dh 6 6 mm Minichannels 200 lm < Dh 6 3 mm
Meso-channels 100 lm < Dh 6 1 mm Microchannels 10 lm < Dh 6 200 lm
Micro-channels 1 lm < Dh 6 100 lm Transitional channels 0.1 lm < Dh 6 10 lm
Dh 6 0.1 lm
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 273

Table 6
Summary of parameters that effect TEG based on literature review.

Heat sink type Author Parameters used. Results

Metal plate [153] Hot side temperature and air speed Compared with passive finned heat sink and finned condenser heat pipe passive metal
plate has lower heat flux
Fin heat sink [117] Fin spacing, fin length and fin frontal area The design method recommended that reducing the length of the heat sink fins by
increasing fin tip. Decreasing TEG number when the heat sink length is below 14.5 mm
caused improvement of power output
Heat pipe [118] Forced air velocity on HPHS The heat pipe with TEG system has the ability of recovering 1.345 kW of waste heat and
yielding 10.39 W of electricity using 8 installed TEGs
Heat pipe [126] Air mass flow rate ratio and air velocity A higher mass flow rate ratio results in a higher amount of heat transfer and higher power
output
Heat pipe [115] Hot side temperature and air speed Heat pipe with finned condenser has best heat flux compared with metal plate and finned
block
Fan-fin heat sink [132] Heat sink surface area and temperature To improve performance of this system, heat sink surface area should expand in an
difference appropriate range and improve cold-side heat transfer capacity
Fan-fin heat sink [133] The thermal resistance of the cold and hot The study showed it was significant to reach a similar enhancement on hot and cold sides
side of heat sink of thermal resistance in order to increase the TEG power output
Fan-fin heat sink [134] Fins number, heat transfer coefficient, fin Enhanced power production due to developed TE materials and optimized heat sink
height and TEG materials height and heat transfer coefficient
Fan-fin heat sink [135] The inlet temperature of hot air, flow rate Some critical heat-transfer parameters on integral TEG system were provided and
of hot air and flow rate of cooling air discussed for further reference
Liquid cooled plate [140] Velocity of fluid, fluid direction and The influence of stagnancy point was determined. The friction along the fluid surface was
software mesh calculated
Liquid cooled plate [141] Hot and cold fluid inlet temperatures, The hot fluid inlet temperature and flow rate both have significant effect on efficiency and
flow rates and the load resistance power output of TEG
Microchannel [150] Flow rates There is a unique coolant flow rate at any DT tegav that makes maximum net-power in the
system
Microchannel [2] Mass flow rates, heat fluxes and a An optimum pumping power is achieved
pressure drop
Microchannel [151] Mass flow and flow conditions Micro-channel offered higher pressure drop, but less heat input to TEG and mass flow rate
is needed to offer the same produced power
Microchannel [152] Channel width, channel height and fin There are specific amounts of channel width and fin thickness that offer maximum power
thickness output in the TEG, for every pumping power

mechanics of the fluid within the micro channel of heat sinks. The t
Rcond ¼ ð45Þ
study of fluid flow and heat transfer can be grouped into two cat- K s WL
egories with respect to the stage of the coolant flowing via them, h  i
namely; two-phase flow and single phase flow [147,148]. The main Ln 1= sin p2  1þW1c =W
Rcont ¼ ð46Þ
f
barriers and challenges faced by micro channels range from leaks
in coolants, poor distribution of the liquid, lack of uniformity in K s WL
the temperature and a large drop in pressure in the direction of h  i
flow which leads to a rise in the requirement of the pumping Ln 1= sin p2  1þW1c =W
Rspread ¼ ð47Þ
f
power. Micro channels are at present being utilized for top of the K s WL
line applications. They do not yet supplant the traditional channels
commercially, principally in view of the high cost connected with 1
specific creation methods required to fabricate micro- and mini- RConv ;fin ¼ ð48Þ
h  n  L  W c  2  h  gf  n  L  Hc
channels [142].
Heat sinks can be classified in relations to the direction of flow
of the fluid inside them: 1
RConv ;c ¼ ð49Þ
h  n  L  Wc
i. Single stack parallel flow heat sink: Consisting of one layer of
parallel channels where in each of the channels the fluid 1
Rbulk ¼ ð50Þ
flows in the same direction [Fig 15(a)]; q  C p  v m  n  Hc  W c
ii. Single stack counter flow heat sink: Consisting of a single Rezenai and Rosendahl [150] studied the power generated
layer of parallel channels where in the adjacent channels based on the pumping power of the coolant in the TEG. The study
the fluid is designed to flow in the opposite ways [Fig 15(b)]; included the optimum rates of coolant flow which influence the
iii. Parallel flow multi-stack heat sinks: Consisting of many lay- maximum TEG power output. The outcomes indicated that there
ers of channels stacked over each other while the coolant is a distinct rate of fluid flow at an average value of DT of the
flows through all channels in the same direction [Fig 15 TEG which constitutes the system’s maximum net-power. There
(c)]; and is an increase in the value of optimum flow with an increase in
iv. Counter flow multi-stack heat sinks: Consisting of multiple the average DT of TEG. The results also show that there is a slow
layers of channels stacked on each other while the coolant increase in the DT between the hot and cold sides proportional
flows in opposite directions with respect to a pair of adjacent to the rate of flow at a constant heat removal. This small difference
stacks [Fig 15(d)]. is due to the little effect decrease based on the thermal resistance
of the heat sink. An experiment and thermal analysis based on TEG
on a parallel micro channel heat sink in addition to their thermal
8.2.3.1. Modelling of micro-channel heat sink. To calculate these reaction was conducted and discussed by Rezania and Rosendahl
resistances, one-dimensional conduction is assumed as shown in [2]. The operating conditions of a real TEG could be solved by the
Fig. 16: three-dimensional principle equations for heat transfer and fluid
274 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

Fig. 15. Diagrams showing rectangular cross-section channels (a) a single stack parallel flow heat sink, (b) a single stack counter flow heat sink, (c) a parallel flow multi-stack
heat sinks and (d) a counter flow multi-stack heat sink.

arrangement of the plenum had an impact on the distribution of

heat transfer in the micro channels. With respect to the arrange-
ment of the micro channels, the rate of mass flow distribution
had a distinct profile in the micro channels at distinct pressure
drops.
Rezania and Rosendahl [151] made a comparison through an
experiment between a traditional heat sink and a micro channel
heat sink in their application to a thermo-electric power generator.
They evaluated the pros and cons of utilizing each type of heat sink
with TEG. The diameter of the micro channel hydraulic is
5.33  104 mm while that of the micro channels is
2.13  103 mm. The heat removed and the pressure drop in the
configuration of the micro heat sink for six distinct rates of mass
flow for turbulent and laminar fluid flow regimes were then
recorded. The thermal and fluid parameters for both turbulent
and laminar regimes were also done through a computational
application of a constant temperature difference between the cold
and hot sides of the TEG.
The results showed that there was a higher pressure drop from
the micro channel heat sink but the heat flow and mass flow rate
across the TEG required for the generation of the same power were
Fig. 16. Thermal resistance network for a three-layered micro-channel stack [149]. lesser for the micro channel heat sink. Investigation and optimiza-
tion of the micro channel heat sink used by the TEG was also made
so as to maximize the cost performance and output power of the
flow in the laminar flow regime with the help of the commercial generic TEG systems [152]. The model utilizes the effective numer-
computational fluid dynamic (CFD) and finite volume method ical test to discuss the optimum size of the system components
(FVM). There was also a study of the thermal performance of the dimensions in two areas of the substrate plate of the heat sink.
flow based on the limitation for maximum temperature for Bi2Te3 With respect to the results, it is clear there is switching of the max-
material alongside the heat sink based on the three fluxes of heat imum output power by the larger substrate plate to the width of
and a vast array of pressure drops. The outcomes indicated that the larger channel, optimum thermo-element length and fin thick-
the increase in the pressure drop of the channels fixed under the ness for maximizing the output power. For all the pumping powers
legs produced a higher portion of the heat applied in relation to discussed, it was found that maximum power of the TEG is deter-
the ones fixed in the heat sink between the two legs. There was mined by a specific value of fin thickness, channel height and chan-
a higher increase in the temperature of the liquid in these channels nel width. There are also requirements based on smaller fill factor,
in relation to the channels located between the two legs in the heat larger channel height and width, as well as lower pumping power
sinks. The distinct rates of mass flow in the channels because of the for the production of maximum cost performance.
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 275

9. Conclusion [7] Gayner C, Kar KK. Recent advances in thermoelectric materials. Prog Mater Sci
2016;83:330–82.
[8] Wu H et al. Advanced electron microscopy for thermoelectric materials. Nano
This paper started with a brief background on the principles and Energy 2015;13:626–50.
theories of TEGs, with their significance and applications on waste- [9] Elsheikh MH et al. A review on thermoelectric renewable energy: principle
parameters that affect their performance. Renew Sustain Energy Rev
heat energy reviewed and discussed. The materials used in the TEG
2014;30:337–55.
device and their impact on the conversion efficiency are also elab- [10] Biswas K et al. High-performance bulk thermoelectrics with all-scale
orated. Towards the end, the paper focuses more on the classifica- hierarchical architectures. Nature 2012;489(7416):414–8.
[11] Pei Y et al. Convergence of electronic bands for high performance bulk
tion of the heat sinks and their application with TEG.
thermoelectrics. Nature 2011;473(7345):66–9.
Studies and research on thermo-electric generation had [12] Hsu KF et al. Cubic AgPbmSbTe2+ m: bulk thermoelectric materials with high
received attention ever since the Seebeck effect was discovered figure of merit. Science 2004;303(5659):818–21.
by Thomas Seebeck in 1821. However, technological development [13] Wu H et al. Broad temperature plateau for thermoelectric figure of merit ZT >
2 in phase-separated PbTe0. 7S0. 3. Nat Commun 2014;5.
and application was slow due to the low thermo-electric conver- [14] Sugar J, Medlin D. Solid-state precipitation of stable and metastable
sion efficiency. Until 1950, the conversion efficiency achieved layered compounds in thermoelectric AgSbTe2. J Mater Sci 2011;46(6):
was less than 5%. Research on TEG was almost dormant from 1668–79.
[15] Poudel B et al. High-thermoelectric performance of nanostructured bismuth
1950 to 1990. It was only after 1990, with major new breakthrough antimony telluride bulk alloys. Science 2008;320(5876):634–8.
and commercially viable renewable energy technologies (such as [16] Wu D et al. Origin of the high performance in GeTe-based thermoelectric
wind turbines and photovoltaic systems), that studies on waste materials upon Bi2Te3 doping. J Am Chem Soc 2014;136(32):11412–9.
[17] Li J et al. High thermoelectric performance of Ge 1  xPbxSe 0.5 Te 0.5 due to
heat recovery technologies received a renewed interest. After (Pb, Se) co-doping. Acta Mater 2014;74:215–23.
1990 s conversion efficiency had reached up to 20%, providing a [18] Plirdpring T et al. Chalcopyrite CuGaTe2: a high-efficiency bulk
new life to a faster rate of technological progress of TEGs in the thermoelectric material. Adv Mater 2012;24(27):3622–6.
[19] Zhang J et al. High-performance pseudocubic thermoelectric materials from
near future.
non-cubic chalcopyrite compounds. Adv Mater 2014;26(23):3848–53.
Two inter-dependent factors on the development of thermo- [20] Lin J et al. Unexpected high-temperature stability of b-Zn4Sb3 opens the door
electric efficiency are expected to enhance future performance of to enhanced thermoelectric performance. J Am Chem Soc 2014;136
(4):1497–504.
TEG; firstly by enhancements in thermo-electric materials through
[21] Snyder GJ et al. Disordered zinc in Zn4Sb3 with phonon-glass and electron-
increasing the figure of merit (ZT) by using novel thermo-electric crystal thermoelectric properties. Nat Mater 2004;3(7):458–63.
materials and secondly, by further improvement of thermal design [22] Joshi G et al. Enhancement in thermoelectric figure-of-merit of an N-type
of heat exchangers through enhancement of heat dissipation using half-heusler compound by the nanocomposite approach. Adv Energy Mater
2011;1(4):643–7.
heat sinks. Increase in TEG efficiency is dependent on good selec- [23] Liu Y et al. Large enhancements of thermopower and carrier mobility in
tion and careful designing of heat sinks. The following considera- quantum dot engineered bulk semiconductors. J Am Chem Soc 2013;135
tions are important when choosing existing heat sinks options or (20):7486–95.
[24] Makongo JP et al. Simultaneous large enhancements in thermopower and
designing new ones. Firstly, the heat sinks applied to the TEG are electrical conductivity of bulk nanostructured half-Heusler alloys. J Am Chem
dependent upon applications of power density. Secondly, passive Soc 2011;133(46):18843–52.
heat sinks are most appropriate because of the inherent efficiency [25] Wang X et al. Compound defects and thermoelectric properties in ternary
CuAgSe-based materials. J Mater Chem A 2015;3(26):13662–70.
of TEG since they lack auxiliary power consumption. Thirdly, other [26] Shi X et al. Multiple-filled skutterudites: high thermoelectric figure of merit
than not having auxiliary power consumption another efficient through separately optimizing electrical and thermal transports. J Am Chem
approach is to use phase change and liquid cold heat sinks. Soc 2011;133(20):7837–46.
[27] Lee Y et al. Contrasting role of antimony and bismuth dopants on the
Fourthly, when heat pipe heat sinks (HPHS) are incorporated as
thermoelectric performance of lead selenide. Nat Commun 2014;5.
heat exchangers, medium temperature range below 300 °C is [28] Rhyee J-S et al. Peierls distortion as a route to high thermoelectric
found to be most suitable. Fifthly, although micro-channel heat performance in In4Se3-&dgr; crystals. Nature 2009;459(7249):965–8.
[29] Zhao L-D et al. Ultralow thermal conductivity and high thermoelectric figure
sinks can significantly improve the heat flow thus further enhanc-
of merit in SnSe crystals. Nature 2014;508(7496):373–7.
ing the electricity generation, in view of the high cost connected [30] Liu R et al. Ternary compound CuInTe 2: a promising thermoelectric material
with specific creation methods required to fabricate the micro- with diamond-like structure. Chem Commun 2012;48(32):3818–20.
channels, they are not yet able to supplant the traditional channels. [31] Liu H et al. Ultrahigh thermoelectric performance by electron and phonon
critical scattering in Cu2Se1-xIx. Adv Mater 2013;25(45):6607–12.
Last but not least, it is helpful to take advantage of the exhaust heat [32] Liu H et al. Copper ion liquid-like thermoelectrics. Nat Mater 2012;11
or cooling sources [154] in the industry, or even in domestic appli- (5):422–5.
cations in order to cool the TEG without extra energy consumption [33] Ibáñez M et al. Cu2ZnGeSe4 nanocrystals: synthesis and thermoelectric
properties. J Am Chem Soc 2012;134(9):4060–3.
for cooling which reduces its conversion efficiency. [34] Zeier WG et al. Phonon scattering through a local anisotropic structural
disorder in the thermoelectric solid solution Cu2Zn1 – x Fe x GeSe4. J Am
Chem Soc 2013;135(2):726–32.
Acknowledgement
[35] Zhao L-D et al. Raising the thermoelectric performance of p-type PbS with
endotaxial nanostructuring and valence-band offset engineering using CdS
The authors would like to thank Universiti Malaysia Pahang for and ZnS. J Am Chem Soc 2012;134(39):16327–36.
providing Research Grant under (RDU150380 & RDU160310). [36] Wu H et al. Strong enhancement of phonon scattering through nanoscale
grains in lead sulfide thermoelectrics. NPG Asia Mater 2014;6(6):e108.
[37] Wu L et al. Origin of phonon glass-electron crystal behavior in thermoelectric
References layered cobaltate. Adv Funct Mater 2013;23(46):5728–36.
[38] Van Nong N et al. Enhancement of the thermoelectric performance of p-type
layered oxide Ca3Co4O9+ d through heavy doping and metallic
[1] Saidur R et al. Technologies to recover exhaust heat from internal combustion
nanoinclusions. Adv Mater 2011;23(21):2484–90.
engines. Renew Sustain Energy Rev 2012;16(8):5649–59.
[39] He Y et al. High thermoelectric performance in non-toxic earth-abundant
[2] Rezania A, Rosendahl L. Thermal effect of a thermoelectric generator on
copper sulfide. Adv Mater 2014;26(23):3974–8.
parallel microchannel heat sink. Energy 2012;37(1):220–7.
[40] Gerlinger M et al. Intratumor heterogeneity and branched evolution revealed
[3] Yazawa K et al. Thermoelectric topping cycles for power plants to eliminate
by multiregion sequencing. N Engl J Med 2012;366(10):883–92.
cooling water consumption. Energy Convers Manage 2014;84:244–52.
[41] Gascoin F et al. Zintl phases as thermoelectric materials: tuned transport
[4] Yazawa K, Koh YR, Shakouri A. Optimization of thermoelectric topping
properties of the compounds CaxYb1–xZn2Sb2. Adv Funct Mater 2005;15
combined steam turbine cycles for energy economy. Appl Energy
(11):1860–4.
2013;109:1–9.
[42] Tan G et al. High thermoelectric performance of p-type SnTe via a synergistic
[5] Hsu C-T et al. Experiments and simulations on low-temperature waste heat
band engineering and nanostructuring approach. J Am Chem Soc 2014;136
harvesting system by thermoelectric power generators. Appl Energy 2011;88
(19):7006–17.
(4):1291–7.
[43] Zhao L-D et al. BiCuSeO oxyselenides: new promising thermoelectric
[6] Liu W et al. Current progress and future challenges in thermoelectric power
materials. Energy Environ Sci 2014;7(9):2900–24.
generation: from materials to devices. Acta Mater 2015;87:357–76.
276 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277

[44] Pan L, Bérardan D, Dragoe N. High thermoelectric properties of n-type [80] Xi H, Luo L, Fraisse G. Development and applications of solar-based
AgBiSe2. J Am Chem Soc 2013;135(13):4914–7. thermoelectric technologies. Renew Sustain Energy Rev 2007;11(5):923–36.
[45] Xiao C et al. High thermoelectric and reversible pnp conduction type [81] Fthenakis V, Kim HC. Life-cycle uses of water in US electricity generation.
switching integrated in dimetal chalcogenide. J Am Chem Soc 2012;134 Renew Sustain Energy Rev 2010;14(7):2039–48.
(44):18460–6. [82] Martín-González M, Caballero-Calero O, Díaz-Chao P. Nanoengineering
[46] Morelli D, Jovovic V, Heremans J. Intrinsically minimal thermal conductivity thermoelectrics for 21st century: energy harvesting and other trends in the
in cubic I V VI 2 semiconductors. Phys Rev Lett 2008;101(3):035901. field. Renew Sustain Energy Rev 2013;24:288–305.
[47] Toberer ES et al. High temperature thermoelectric efficiency in Ba 8 Ga 16 Ge [83] Shu G et al. A review of waste heat recovery on two-stroke IC engine aboard
30. Phys Rev B 2008;77(7):075203. ships. Renew Sustain Energy Rev 2013;19:385–401.
[48] He H et al. Synthesis, structural characterization, and physical properties of [84] Tchanche BF et al. Low-grade heat conversion into power using organic
the type-I clathrates A 8Zn18As28 (A = K, Rb, Cs) and Cs8Cd18As28. Chem Rankine cycles – a review of various applications. Renew Sustain Energy Rev
Mater 2012;24(18):3596–603. 2011;15(8):3963–79.
[49] Liu W et al. Convergence of conduction bands as a means of enhancing [85] Wang T et al. A review of researches on thermal exhaust heat recovery with
thermoelectric performance of n-type Mg 2 Si 1 x Sn x solid solutions. Phys Rankine cycle. Renew Sustain Energy Rev 2011;15(6):2862–71.
Rev Lett 2012;108(16):166601. [86] Dai D, Zhou Y, Liu J. Liquid metal based thermoelectric generation system for
[50] Xiao C et al. Superionic phase transition in silver chalcogenide nanocrystals waste heat recovery. Renew Energy 2011;36(12):3530–6.
realizing optimized thermoelectric performance. J Am Chem Soc 2012;134 [87] Riffat SB, Ma X. Thermoelectrics: a review of present and potential
(9):4287–93. applications. Appl Therm Eng 2003;23(8):913–35.
[51] Rowe D, Shukla V, Savvides N. Phonon scattering at grain boundaries in [88] Tie SF, Tan CW. A review of energy sources and energy management system
heavily doped fine-grained silicon–germanium alloys; 1981. in electric vehicles. Renew Sustain Energy Rev 2013;20:82–102.
[52] Yu B et al. Enhancement of thermoelectric properties by modulation-doping [89] Ullah K et al. A review of solar thermal refrigeration and cooling methods.
in silicon germanium alloy nanocomposites. Nano Lett 2012;12(4):2077–82. Renew Sustain Energy Rev 2013;24:499–513.
[53] Klemeš JJ, Varbanov PS. Heat integration including heat exchangers, [90] Fleurial J-P. Thermoelectric power generation materials: technology and
combined heat and power, heat pumps, separation processes and process application opportunities. JOM 2009;61(4):79–85.
control. Appl Therm Eng 2012;43:1–6. [91] Astrain D, Martínez Á. Heat exchangers for thermoelectric devices. INTECH
[54] Bendig M, Maréchal F, Favrat D. Defining ‘‘Waste Heat” for industrial Open Access Publisher; 2012.
processes. Appl Therm Eng 2013;61(1):134–42. [92] Heremans JP et al. Enhancement of thermoelectric efficiency in PbTe by
[55] De Beer J, Worrell E, Blok K. Long-term energy-efficiency improvements in distortion of the electronic density of states. Science 2008;321(5888):554–7.
the paper and board industry. Energy 1998;23(1):21–42. [93] Lertsatitthanakorn C. Electrical performance analysis and economic
[56] Johnson I et al. Waste heat recovery: technology and opportunities in US evaluation of combined biomass cook stove thermoelectric (BITE)
industry. US Department of Energy, Office of Energy Efficiency and generator. Bioresour Technol 2007;98(8):1670–4.
Renewable Energy, Industrial Technologies Program, 2008. [94] Snyder J. Small thermoelectric generators. The Electrochemical Society; 2008.
[57] Yang J. Potential applications of thermoelectric waste heat recovery in the p. 54–6. doi: h ttp.
automotive industry. In: ICT 2005. 24th international conference on [95] Twaha S et al. A comprehensive review of thermoelectric technology:
thermoelectrics, 2005. IEEE; 2005. materials, applications, modelling and performance improvement. Renew
[58] Gao H et al. Development of stove-powered thermoelectric generators: a Sustain Energy Rev 2016;65:698–726.
review. Appl Therm Eng 2016;96:297–310. [96] Vullers R et al. Micropower energy harvesting. Solid-State Electr 2009;53
[59] Zheng X et al. A review of thermoelectrics research – recent developments (7):684–93.
and potentials for sustainable and renewable energy applications. Renew [97] Commission E.Critical raw materials for the EU. Report of the Ad-hoc Working
Sustain Energy Rev 2014;32:486–503. Group on defining critical raw materials. Ad-hoc Working Group: July, 2010;
[60] He W et al. Recent development and application of thermoelectric generator 2010. p. 84.
and cooler. Appl Energy 2015;143:1–25. [98] Liu X et al. An energy-harvesting system using thermoelectric power
[61] Patil D, Arakerimath DR. A review of thermoelectric generator for waste heat generation for automotive application. Int J Electr Power Energy Syst
recovery from engine exhaust. Int J Res Aeronaut Mech Eng 2013;1:1–9. 2015;67:510–6.
[62] Cai K et al. The effect of titanium diboride addition on the thermoelectric [99] LaGrandeur J et al. Automotive waste heat conversion to electric power using
properties of b-FeSi 2 semiconductors. Solid State Commun 2004;131 skutterudite, TAGS, PbTe and BiTe. In: 2006 25th international conference on
(5):325–9. thermoelectrics. IEEE; 2006.
[63] Handbook T. Macro to nano. In: Rowe DM, editor. CRC Press, Taylor& Francis; [100] Yang J. Potential applications of thermoelectric waste heat recovery in the
2006. automotive industry. In: ICT 2005. 24th international conference on
[64] Chasmar R, Stratton R. The thermoelectric figure of merit and its relation to thermoelectrics, 2005. IEEE; 2005.
thermoelectric generators. Int J Electron 1959;7(1):52–72. [101] Ismail BI, Ahmed WH. Thermoelectric power generation using waste-heat
[65] Gao X et al. Rational design of high-efficiency thermoelectric materials with energy as an alternative green technology. Recent Patents Electr Electron Eng
low band gap conductive polymers. Comput Mater Sci 2006;36(1):49–53. (Formerly Recent Patents Electr Eng) 2009;2(1):27–39.
[66] Goldsmid HJ. Bismuth telluride and its alloys as materials for thermoelectric [102] Yodovard P, Khedari J, Hirunlabh J. The potential of waste heat thermoelectric
generation. Materials 2014;7(4):2577–92. power generation from diesel cycle and gas turbine cogeneration plants.
[67] Tani J-I, Kido H. Thermoelectric properties of Bi-doped Mg 2 Si Energy Sources 2001;23(3):213–24.
semiconductors. Physica B 2005;364(1):218–24. [103] LeBlanc S. Thermoelectric generators: linking material properties and
[68] Sootsman JR, Chung DY, Kanatzidis MG. New and old concepts in systems engineering for waste heat recovery applications. Sustain Mater
thermoelectric materials. Angew Chem Int Ed 2009;48(46):8616–39. Technol 2014;1:26–35.
[69] Ovik R et al. A review on nanostructures of high-temperature thermoelectric [104] LaGrandeur J, Crane D, Eder A. Vehicle fuel economy improvement through
materials for waste heat recovery. Renew Sustain Energy Rev thermoelectric waste heat recovery. In: DEER conference; 2005.
2016;64:635–59. [105] Hendricks TJ et al. New perspectives in thermoelectric energy recovery
[70] Aswal DK, Basu R, Singh A. Key issues in development of thermoelectric system design optimization. J Electron Mater 2013;42(7):1725–36.
power generators: high figure-of-merit materials and their highly conducting [106] Hendricks T, Choate WT. Engineering scoping study of thermoelectric
interfaces with metallic interconnects. Energy Convers Manage generator systems for industrial waste heat recovery. US Dept Energy
2016;114:50–67. 2006;20:6.
[71] Dughaish Z. Lead telluride as a thermoelectric material for thermoelectric [107] Johnson I, Choate WT, Davidson A. Waste Heat Recovery. Technology and
power generation. Physica B 2002;322(1):205–23. Opportunities in US Industry. Laurel (MD, United States): BCS, Inc.; 2008.
[72] Majumdar A. Thermoelectricity in semiconductor nanostructures. Science [108] Deng Y et al. Enhanced performance of solar-driven photovoltaic–
2004;303(5659):777–8. thermoelectric hybrid system in an integrated design. Sol Energy
[73] Orr B et al. A review of car waste heat recovery systems utilising 2013;88:182–91.
thermoelectric generators and heat pipes. Appl Therm Eng 2016;101:490–5. [109] Kraemer D et al. High-performance flat-panel solar thermoelectric generators
[74] Zhu Y et al. Influence of Dy/Bi dual doping on thermoelectric performance of with high thermal concentration. Nat Mater 2011;10(7):532–8.
CaMnO3 ceramics. Mater Chem Phys 2014;144(3):385–9. [110] Miljkovic N, Wang EN. Modeling and optimization of hybrid solar
[75] Yasukawa M et al. High-temperature thermoelectric properties of La-doped thermoelectric systems with thermosyphons. Sol Energy 2011;85
BaSnO3 ceramics. Mater Sci Eng, B 2010;173(1):29–32. (11):2843–55.
[76] Kalkan N, Young E, Celiktas A. Solar thermal air conditioning technology [111] Zhang M et al. Efficient, low-cost solar thermoelectric cogenerators
reducing the footprint of solar thermal air conditioning. Renew Sustain comprising evacuated tubular solar collectors and thermoelectric modules.
Energy Rev 2012;16(8):6352–83. Appl Energy 2013;109:51–9.
[77] Afshar O et al. A review of thermodynamics and heat transfer in solar [112] Kim S, So S, Ki H. Controlling thermal deformation using a heat sink in laser
refrigeration system. Renew Sustain Energy Rev 2012;16(8):5639–48. transformation hardening of steel sheets. J Mater Process Technol
[78] Thirugnanasambandam M, Iniyan S, Goic R. A review of solar thermal 2015;216:455–62.
technologies. Renew Sustain Energy Rev 2010;14(1):312–22. [113] Lee HS. Thermal design: heat sinks, thermoelectrics, heat pipes, compact heat
[79] Omer AM. Focus on low carbon technologies: the positive solution. Renew exchangers, and solar cells. John Wiley & Sons; 2010.
Sustain Energy Rev 2008;12(9):2331–57. [114] Guide HSS. SO8-FL; 2011.
 A. Elghool et al. / Energy Conversion and Management 134 (2017) 260–277 277

[115] Date A et al. Theoretical and experimental estimation of limiting input heat [136] Maddipati UR, Rajendran P, Laxminarayana P. Thermal design and analysis of
flux for thermoelectric power generators with passive cooling. Sol Energy cold plate with various proportions of ethyl glycol water solutions.
2015;111:201–17. [137] Song S, Au V, Moran KP. Constriction/spreading resistance model for
[116] Rowe DM. CRC handbook of thermoelectrics. CRC Press; 1995. electronics packaging. In: Proceedings of the 4th ASME/JSME thermal
[117] Chen W-H. Design of heat sink for improving the performance of engineering joint conference; 1995.
thermoelectric generator. Energy 2012;39:236–45. [138] Cengel YA. Heat transfer a practical approach; 2003.
[118] Remeli MF et al. Simultaneous power generation and heat recovery using a [139] Rajendran P, Ravindra MU, Prasad C. Studies on effect of spreading resistance
heat pipe assisted thermoelectric generator system. Energy Convers Manage in design of variable heat flux with identical heat sink. Int J Electron Eng Res
2015;91:110–9. 2010;2(3):399–408.
[119] Faghri A. Heat pipe science and technology. Global Digital Press; 1995. [140] Suzuki RO et al. Effects of fluid directions on heat exchange in thermoelectric
[120] Faghri A. Review and advances in heat pipe science and technology. J Heat generators. J Electron Mater 2012;41(6):1766–70.
Transf 2012;134(12):123001. [141] Niu X, Yu J, Wang S. Experimental study on low-temperature waste heat
[121] Heat Pipe Wick Structures. Available from: <http://www.pscmf.com/product/ thermoelectric generator. J Power Sources 2009;188(2):621–6.
showproduct>. [142] Dixit T, Ghosh I. Review of micro-and mini-channel heat sinks and heat
[122] Shabgard H et al. Heat pipe heat exchangers and heat sinks: opportunities, exchangers for single phase fluids. Renew Sustain Energy Rev
challenges, applications, analysis, and state of the art. Int J Heat Mass Transf 2015;41:1298–311.
2015;89:138–58. [143] Tuckerman DB, Pease R. High-performance heat sinking for VLSI. IEEE
[123] Rosenfeld JH et al. An overview of long duration sodium heat pipe tests; Electron Device Lett 1981;2(5):126–9.
2004. [144] Keyes RW. Physical limits in digital electronics. Proc IEEE 1975;63
[124] Gholami A, Ahmadi M, Bahrami M. A new analytical approach for dynamic (5):740–67.
modeling of passive multicomponent cooling systems. J Electron Packag [145] Mehendale S, Jacobi A, Shah R. Fluid flow and heat transfer at micro-and
2014;136(3):031010. meso-scales with application to heat exchanger design. Appl Mech Rev
[125] Date A et al. Theoretical and experimental study on heat pipe cooled 2000;53(7):175–93.
thermoelectric generators with water heating using concentrated solar [146] Kandlikar SG, Grande WJ. Evolution of microchannel flow passages-
thermal energy. Sol Energy 2014;105:656–68. thermohydraulic performance. In: Technology and society and engineering
[126] Remeli MF, Singh B, Akbarzadeh A. Passive power generation and heat business management–2002: presented at the 2002 ASME international
recovery from waste heat. Adv Mater Res 2015;1113. mechanical engineering congress and exposition, November 17–22, 2002,
[127] Orr B et al. Electricity generation from an exhaust heat recovery system New Orleans, Louisiana. 2002. American Society of Mechanical Engineers.
utilising thermoelectric cells and heat pipes. Appl Therm Eng 2014;73 [147] Hassan I, Phutthavong P, Abdelgawad M. Microchannel heat sinks: an
(1):588–97. overview of the state-of-the-art. Microscale Thermophys Eng 2004;8
[128] Chingulpitak S, Wongwises S. A review of the effect of flow directions and (3):183–205.
behaviors on the thermal performance of conventional heat sinks. Int J Heat [148] Ganesh Hegde, P., Microchannel heat sinks for cooling high heat flux
Mass Transf 2015;81:10–8. electronic devices-analysis with single and two phase flows, USM; 2006.
[129] Active Heat Sink. [web page] [cited 2016 7/4/2016]; Available from: <https:// [149] Wei X, Joshi Y. Optimization study of stacked micro-channel heat sinks for
www.alphanovatech.com>. micro-electronic cooling. IEEE Trans Compon Packag Technol 2003;26
[130] Zheng X et al. Experimental study of a domestic thermoelectric cogeneration (1):55–61.
system. Appl Therm Eng 2014;62(1):69–79. [150] Rezania A, Rosendahl L, Andreasen SJ. Experimental investigation of
[131] Teertstra P, Yovanovich M, Culham J. Analytical forced convection modeling thermoelectric power generation versus coolant pumping power in a
of plate fin heat sinks. J Electron Manuf 2000;10(04):253–61. microchannel heat sink. Int Commun Heat Mass Transf 2012;39(8):1054–8.
[132] Gou X, Xiao H, Yang S. Modeling, experimental study and optimization on [151] Rezania A, Rosendahl L. Evaluating thermoelectric power generation device
low-temperature waste heat thermoelectric generator system. Appl Energy performance using a rectangular microchannel heat sink. J Electron Mater
2010;87(10):3131–6. 2011;40(5):481–8.
[133] Astrain D et al. Study of the influence of heat exchangers’ thermal resistances [152] Rezania A et al. Co-optimized design of microchannel heat exchangers and
on a thermoelectric generation system. Energy 2010;35(2):602–10. thermoelectric generators. Int J Therm Sci 2013;72:73–81.
[134] Barma M et al. Estimation of thermoelectric power generation by recovering [153] Heat Pipe Wick Structures [web page] [cited 2016 7/4/2016]; Available from:
waste heat from biomass fired thermal oil heater. Energy Convers Manage <http://www.pscmf.com/product/showproduct>.
2015;98:303–13. [154] Soni SK, Pandey M, Bartaria VN. Hybrid ground coupled heat exchanger
[135] Tzeng S-C, Jeng T-M, Lin Y-L. Parametric study of heat-transfer design on the systems for space heating/cooling applications: a review. Renew Sustain
thermoelectric generator system. Int Commun Heat Mass Transf Energy Rev 2016;60:724–38.
2014;52:97–105.