Applied Thermal Engineering 102 (2016) 1045–1058

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Research Paper

Heat transfer and performance analysis of thermoelectric stoves

Yousef S.H. Najjar ⇑, Musaab M. Kseibi
Mechanical Engineering Department, Jordan University of Science and Technology, Irbid, Jordan

h i g h l i g h t s

Design and testing of a thermo electric stove.

Three biofuels namely: wood, peat and manure are used.

Heat transfer analysis is detailed.

Resulting thermoelectric energy for vital purposes in remote poor regions.

Evaluation of performance of the stove subcomponents.

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

Article history: Access to electricity is one of the important challenges for remote poor regions of the world. Adding TEG
Received 30 December 2015 (thermoelectric generators) to stoves can provide electricity for the basic benefits such as: operating
Accepted 22 March 2016 radio, light, phones, medical instruments and other small electronic devices. Heat transfer analysis of a
Available online 6 April 2016
multi-purpose stove coupled with 12 TEG modules is presented. This analysis comprises a well aerody-
namically designed combustor, finned TEG base plate, cooker and water heater beside the outer surface
Keywords: for space heating. Heat transfer analysis was also carried out for all the subcomponents of the stove, and
Multi-task stoves
performance predicted against the experimental results.
Thermo electric generator
Heat Transfer analysis
It was found that the maximum power obtained is about 7.88 W using wood, manure or peat with an
Thermo economic analysis average overall efficiency of the stove about 60%.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction After carrying a detailed review of many types of stoves [2–7], it

was decided to design a multi-task stove to help the people living
Twenty percent of worldwide population (1.3 billion people) in remote poor regions and deprived of electricity. This JUST
are still without access to electricity most of them whom live in (Jordan University of Science and Technology) stove serves the pur-
developing countries [1]. Providing a minimum amount of electric- poses of space heating, cooking and water heating besides produc-
ity that operates their basic needs such as light, radio, telephone ing electric energy by using TEG. The latter will serve different
and some medical instruments, improves the quality of their lives aspects of life including: lighting, charging the cellular phone, lis-
to meet the basic needs of population. TEG coupled to the stove can tening to the radio and operates some medical instruments and
be a very interesting option to provide such amount of electricity. much more.
TEG is a device that harvests waste energy and converts it It requires almost no maintenance since there are no moving
partially to useful power. It operates on a fundamental principle parts. The TEG also works day and night even in cloudy or rainy
termed the Seebeck effect. It states that when a temperature gradi- weather unlike solar panels.
ent is established between two different metals or semiconductors, Hence the objective of this work was carrying out the design
a corresponding voltage gradient is induced which causes a contin- and heat transfer performance analysis over a wide range of fuels
uous current to flow through the semiconductors when they form and operating conditions. The design covered the whole compo-
a complete loop. nents including the combustor with special aerothermal design
[8–12]. The fins of both TEG and cooker; TEG base and heat sink
and the cooker were designed to give optimum spacing and perfor-
mance. Water heater was designed with a copper tube exposed
⇑ Corresponding author.
to the exhaust gases to recover most of the remained energy.
E-mail addresses: [email protected] (Y.S.H. Najjar), [email protected]
To simulate the real circumstances of those living in remote areas,
(M.M. Kseibi).

http://dx.doi.org/10.1016/j.applthermaleng.2016.03.114
1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.
1046 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058

Nomenclature

Symbols Tc cold side temperature of the thermoelectric model, K

A surface area, m2 Tg gas temperature, K
Cpg gas specific heat at constant pressure, kJ/kg K TH hot side temperature of the thermoelectric model, K
Cpw water specific heat at constant pressure, kJ/kg K Ts stove body surface temperature, K
F s-surr total view factor between the walls and the base of the Tw water temperature, K
channel and its surrounding T stack stack surface temperature, K
h heat transfer coefficient, W/m2 K T thickness of the stove wall, mm
H fin height U pn thermal conductance of thermoelectric generator, W/K
I electric current, A V oc open circuit voltage, V
JUST Jordan University of Science and Technology
K thermal conductivity, W/m K Greek letters
L fin length, cm E emissivity
LHV lower heating value of the fuel, kJ/kg R Stefan Boltzmann constant 5.67 ⁄ 108 W/m2 K4
m_a mass flow rate of the air, kg/s H efficiency
m_f mass flow rate of fuel, kg/s
m_g mass flow rate of the gas, kg/s Subscripts
m_w mass flow rate of water, kg/s 1–5 position of the gas path along the stove
N number of thermoelectric models A air
P electric power, W Al aluminum
Qa sensible heat in the air, W B base
Qc heat rejected to the heat sink C cold
Q ch radiation heat transfer rate from U-shaped channel, W Cc combustion chamber
Q cook heat absorbed by the cooker plate, W conv convection
Qf sensible heat in the burned fuel, W Cw ceramic wafer
Qg sensible heat in the gas, W F fuel
Q ga heat transfer from the gas to the surrounding air for G has
space heating, W H hot
Q gw heat transfer from the gas to the water inside the coil, W Hs heat sink
Q g4—5 available energy after the cooker, W I inside
QH heat absorbed by the TEG fins, W Ip interface plate
QLoss heat loss, W max maximum
R thermal resistance, K/W ms metallic solder
Ri internal resistance of the TEG model, ohm o outside
RL load resistance, X oc open circuit
Spn Seebeck coefficient, V/K rad radiation
TEG thermoelectric generator Stoich. stoichiometric
Ta ambient air temperature, K W water

different fuels including wood, manure and peat were utilized. The output power of the TEG strongly depends on the temper-
Heat transfer analysis was especially considered with JUST stove ature difference between the hot and cold sides of the model. It
as a case study. was obvious that the use of water as a cooling medium gives the
highest power because of the higher thermal conductivity of the
water and the cold side temperature of the model will not exceed
2. Experimental facility the boiling temperature of the water which allows us to elevate the
hot side temperature as much as the model can take.
This stove has an improved combustor aerodynamic design and Thermoelectric generator mode:
contains 12 TEG modules for electricity generation, in addition to The TEG models are sandwiched between 2 sets of fins, the hot
cooker with specially-designed finned, water heater compact inner one in the way of the hot gas and colder one on the side of the
surfaces. It has a compact outside surface, enough for space heating ambient (Figs. 2–4) (see Table 1).
Fig. 1.
The heat from the combustion gases was partially recovered by 3. Heat transfer analysis
installing fins on both the hot side of the TEG and the cooker. These
fins act as a store of energy and they guide the combustion prod- Assumptions:
ucts along the required path from the combustion chamber, base
of the TEG, fins of the cooker, water heater towards the exit stack. 1. Air is Ideal gas.
To maintain the design healthy, this stove is designed so that 2. Gas flow is a steady flow process.
the user is not exposed to exhaust gases at all. Three types of fuel 3. Excess air (30%).
namely: wood, animal dung (manure) and peat, were tested. They 4. Combustion efficiency inside the combustor of the stove
represent the fuels which are the most available and commonly (gc ¼ 60%) [13].
used in the remote areas around the world. 5. Fins surface temperature is steady with time.
 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058 1047

Plate forces the gas

to flow through the
stove components

Fig. 1. Cross sectional view of the JUST stove. Where T g1 is the gas temperature at position 1, T g2 is the gas temperature at position 2, T g3 is the gas temperature at position 3,
T g4 is the gas temperature at position 4, T c1 is the thermocouple at position 1, T c2 is the thermocouple at position 2, T c3 is the thermocouple at position 3, T c4 is the
thermocouple at position 4, T c5 is the thermocouple at position 5, A1 is the combustor zone, A2 is the after combustor zone, A3 is the after TEG fins zone, A4 is the after cooker
zone, and A5 is the stack.

3.1. Combustor

Energy balance of the combustor (see Fig. 5)

Q a þ gc Q f  Q i;1 ¼ Q g;1 ð1Þ

Q a is the sensible heat in the air is very small and often neglected.
_ f  LHV
Qf ¼ m ð2Þ

_ g Cpg ðT g;1  T a Þ
Q g;1 ¼ m ð3Þ

Heat losses from the combustor body can be calculated by

applying the energy balance on the combustor wall (see Fig. 6):
ðqconv þ qrad Þinside ¼ ðqconv þ qrad Þoutside ð4Þ
Heat transfer to wall inside

Q i;1 ¼ eg rA1 T 4g;1 þ hi A1 ðT g;1  T s;1 Þ ð5Þ

Fig. 2. Top view of the TEG assembly.
Heat transfer from wall outside

6. In the TEG assembly; the only path of heat transfer from one Q O;1 ¼ ho A1 ðT s;1  T a Þ þ erA1 ðT 4s;1  T 4a Þ ð6Þ
side of the pellet to the other is through the TEG material. where A1 is the surface area of the combustion chamber and Ts1 is
7. The thickness of the stove wall is neglected in the calculations the stove wall temperature inside and outside. Assumed to be equal
of the stove ðT s;i ¼ T s;o ¼ T s Þ. to the small thickness.
1048 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058

Fig. 3. Top view of the TEG module.

Fig. 4. Side view of the TEG module.



Table 1 air _a
m
¼ ¼ 5:1 for wood ½14 ð8Þ
TEG module characteristics. fuel Stoich: _ f Stoich:
m
Model characteristics Symbol Value Unit
Assuming 30% excess air
Maximum power P 3.7 W
Load resistance Rl 6.8 ohm _a
m
¼ 5:1  1:3 ¼ 6:63
Internal resistance Ri 6.8 ohm _f
m
Number of semiconducting pairs N 12
Thermal conductance Up-n 0.85 K/W This value will be used in the calculation.
Size of the TEG model 40  40  4 mm

3.1.1. After combustor zone

After the combustor, the temperature of the gas (T g;1 ) drops to
Tg2 due to the heat transfer to the stove wall before the TEG fins.
Mass balance The gas temperature in the second region A2, after the combustor,
_ aþm
_g¼m _f can be calculated by rewriting Eqs. (1), (5), and (6) by replacing A1
m ð7Þ
by (A1 þ A2 ) Qi,1 by Qi,2 and Qg1 by Qg2 as following:

Fig. 5. Energy balance on the combustor.

 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058 1049

Also, QH can be written in terms of the thermal resistances as

follows:
T TEG;fin  T H
QH ¼ ð15Þ
Rfin;b þ Ri:p þ Rc:w þ Rms

tb
Rfin;b ¼ ð16Þ
kcast:iron Ab

t i:p
Ri:p ¼ ð17Þ
kAl Ai:p

t c:w
Rc:w ¼ ð18Þ
Fig. 6. Combustor wall.
kceramic Ac:w

t ms
Rms ¼ ð19Þ
Q a þ gc Q f  Q i;2 ¼ Q g;2 ð9Þ kAl Ams
where:
Q i;2 ¼ eg rðA1 þ A2 ÞT 4g;2 þ hi ðA1 þ A2 ÞðT g;2  T s;2 Þ ð10Þ
Rfin,b: Thermal resistivity of the fin base.
Q O;2 ¼ ho ðA1 þ A2 ÞðT s;2  T a Þ þ erðA1 þ A2 ÞðT 4s;2  T 4a Þ ð11Þ Ri,p: Thermal resistivity of the fin interface plate.
Rc.w: Thermal resistivity of the fin ceramic water of the TEG unit.
where: Rm,s: Thermal resistivity of the metallic solders of TEG unit.

Q i;2 : The energy lost from the total stove walls before the TEG
At the cold side (outside the stove):
fins, kJ.
The heat dissipated by the heat sink is a combination of two
A2 : The surface area of the stove before the TEG fins.
modes convection and radiation.
ðA1 þ A2 Þ is used to get a correct value of the gas temperature
(Tg,2) at position (2) we have to subtract all the heat lost from Q c ¼ Q c;rad þ Q c;conv ð20Þ
the total heat (heating value) before that position.
The heat dissipated by radiation can be calculated using the fol-
Hence, the new temperature obtained (T g;2 ) will be nearly equal lowing correlations [15–18]:
 
to the temperature measured experimentally by the first thermo- Q c;rad ¼ nf  1 Q ch
couple (T c;1 ).   
þ nf tðL þ 2HÞ þ 2HL þ 2t b ðL þ wÞ re T 4HS  T 4a ð21Þ
3.2. TEG assembly

rðS þ 2HÞL T 4HS  T 4a
At the hot side (inside the stove): Q ch ¼ ð22Þ
ð1  e=eÞ þ ð1=F s-surr Þ
Heat absorbed by the inside fins is a combination of two modes
convection and radiation (Fig. 7) h  i
 1 þ L2 0:5  1
2H
Q H ¼ Q H;rad þ Q H;conv ð12Þ F s-surr ¼ 1  ð23Þ
 L þ ð1 þ L2 Þ0:5  1
2H
The heat absorbed by radiation can be calculated using the fol-
lowing relations:  ¼ H=S
H ð24Þ
Q H;rad ¼ eg r Afins T 4g2 ð13Þ
L ¼ L=S ð25Þ
The heat absorbed by convection can be calculated using the
where:
following relation:
   
Q H;conv ¼ hi Aunfin T g2  T fin þ gfin hi Afin T g2  T fin ð14Þ nf : Number of fins.
Q ch : Radiation heat transfer rate from U-shaped channel.
tb : Base thickness.
F s-surr : Total view factor between the walls and the base of the
channel and its surrounding.
 Normalized fin height mm/mm.
H:
L: Normalized fin length mm/mm.
H: Fin height.
L: Fin length.
S: Spacing between fins.

The correlations (21 through 25) are used to simplify the calcu-
lation of the view factor from the heat sink-channel which can be
extremely difficult using normal method.
The heat dissipated by convection:
1050 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058

Fig. 7. The assembly of the TEG.

Q c;conv ¼ ho Aunfin ðT HS  T a Þ þ gfin ho Afin ðT Hs  T a Þ ð26Þ U pn ðT H  T c Þ: Heat absorbed/dissipated by conduction from the
Hot/cold junction.
Also; Qc can be written in terms of thermal resistance as
U pn : Overall heat transfer coefficient W/K.
follows:
ISpn T H : Heat due to Peltier effect.
T c  T HS Spn : Seeback coefficient V/K.
Qc ¼ ð27Þ I2 Ri =2: Resistive heating (reduces the efficiency of the model).
Rs þ RC;w þ RHS;b
N: number of TEG models used.
t HS:b
RHS;b ¼ ð28Þ 3.2.2. Electric part
kAl AHS;b
The maximum power transfer to the load when the load resis-
Thermoelectric generator model (see Fig. 8): tance equals the internal resistance of the model thus (see Fig. 9);
I ¼ V oc =Ri þ RL ¼ V oc =2Ri ð32Þ
3.2.1. Heat part
V oc ¼ Spn ðT H  T c Þ ð33Þ


Ri S2pn ðT H  T c Þ2
Q H ¼ N U pn ðT H  T c Þ þ ISpn T H  I2 ð29Þ Pmax ¼ N  I2 RL ¼ N  ð34Þ
2 4RL


Ri 3.2.3. After TEG zone
Q c ¼ N U pn ðT H  T c Þ þ ISpn T c þ I2 ð30Þ
2 After the TEG assembly, the temperature of the gas (T g;2 ) drop
due to the heat lost through the stove wall after the TEG fins.
P ¼ Q H  Q c ¼ ISpn ðT H  T c Þ  I2 Ri ð31Þ The gas temperature after the TEG fins and before the cooker fins
can be calculated by using Eqs. (1), (5), and (6) by replacing A1
where: by (A1 þ A2 þ A3 ) and accounting for the energy absorbed by the
TEG fins as following:
Q a þ gc Q f  Q loss;3  Q H ¼ Q g;3 ð35Þ
Heat source

Electrical insulator

P N Electrical conductor

Heat sink

Current

Load resistance

Fig. 8. Single p–n pair of the TEG model. Fig. 9. Electric circuit of the TEG module.
 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058 1051

Q i;3 ¼ eg rðA1 þ A2 þ A3 ÞT 4g;3 þ hi ðA1 þ A2 þ A3 ÞðT g;3  T s;3 Þ ð36Þ Q a þ gc Q f  Q loss;4  Q H  Q cook ¼ Q g;4 ð45Þ

Q i;3 ¼ ho ðA1 þ A2 þ A3 ÞðT s;3  T a Þ þ erðA1 þ A2 þ A3 ÞðT 4s;3  T 4a Þ Q i;4 ¼ eg rðA1 þ A2 þ A3 þ A4 ÞT 4g;4
ð37Þ þ hi ðA1 þ A2 þ A3 þ A4 ÞðT g;4  T s;4 Þ ð46Þ

where;
Q i;4 ¼ ho ðA1 þ A2 þ A3 þ A4 ÞðT s;4  T a Þ
Q i;3 : The energy lost from the total stove walls before the þ erðA1 þ A2 þ A3 þ A4 ÞðT 4s;4  T 4a Þ ð47Þ
cooker.
A3 : The surface area of the stove wall after the TEG fins and where:
before the cooker fins.
Q i;4 : The energy lost from the total stove walls before the water
3.3. Cooker heater.
A4 : The surface area of the stove wall after the cooker fins and
The heat absorbed by the cooking model fins from combustion before the water heater (see Fig. 11).
gases is a combination of two modes: convection and radiation
(Fig. 10). 3.4. Water heater (Fig. 12)
The heat absorbed by radiation can be calculated using the fol-
lowing correlations:
Q cook ¼ Q cook;rad þ Q cook;conv ð38Þ Q g4—5 ¼ Q gw þ Q ga ð48Þ

Q cook;rad ¼ eg rAfins T 4g3 ð39Þ _ w Cpw ðT w;i  T w;o Þ

Q gw ¼ m ð49Þ

 
Q cook;conv ¼ hi Aunfin T g3  T Cook;fin þ gfin hi Afin ðT g3  T Cook;fin Þ ð40Þ Q ga ¼ ho A5 ðT stack  T a Þ þ erA5 ðT 4stack  T 4a Þ ð50Þ
Also;
Q ga ¼ eg rA5 T 4g;4 þ hi A5 ðT g;4  T stack Þ ð51Þ
T Cook;fin  T cook;p
Q cook ¼ ð41Þ
Rb _ f gc Q f Þ  Q H  Q cook  Q i;4
Q g4—5 ¼ ðm ð52Þ
T cook;p  T a
Q cook ¼ ð42Þ where:
Ac;p ðho þ hr;o Þ

  Q g4—5 : Available energy after the cooker, W.

hr;o ¼ re T cook;p þ T a T 2cook;p þ T 2a ð43Þ
Q gw : Heat transferred to the water inside the coil, W.
Q ga : Heat lost through the stack wall to the surrounding air, W.
tb m_ w : Mass flow rate of water, kg/s.
Rb ¼ ð44Þ
k b Ab Cpw : Specific heat of water at constant pressure, kJ/kg K.
A5 : Stack surface area surrounding the water heater copper coil,
3.3.1. After cooker zone m2.
After the cooker, the temperature of the gas (T g;3 ) drop due to
the heat lost through the stove wall after the cooker fins. The gas 3.5. Space heating from stove body
temperature after the cooker fins and before the stack entrance
can be calculated by using Eqs. (1), (5), and (6) with replacing A1
by (A1 þ A2 þ A3 þ A4 ) and accounting for the energy absorbed by Q heating ¼ Q i;4 þ Q ga ð53Þ
the TEG fins and the cooker fins as follows:

Fig. 10. Cooker fins.

1052 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058

Q cooking : Heat absorbed by the cooking model fins, W.

Q i;4 : Total heat loss through all the stove walls except the stack,
W.
Q gw : Heat absorbed by the water heater coils, W.
Q ga : Heat lost through the stack wall to the surrounding air, W.
ðm_ f  LHVÞ: Heat input, W.

4. Heat transfer analysis with JUST stove as a case study

4.1. Combustor

_a
m m_a
¼ ¼ 6:63 ð8Þ
_
mf 3:52  104

_ a ¼ 0:00233
m

_ g ¼ 0:00233 þ 3:52  104 2:68  103

m ð7Þ
From here onward, each equation in the previous mode is
rewritten with the following line, the values are compensated to
get the results (see Table 2):

Q a þ gc Q f  Q i;1 ¼ Q g;1 ð1Þ

_ f gc LHV  Q i;1 ¼ m
m _ g Cpg ðT g;1  T a Þ
Fig. 11. Water heater.  
0:000352  0:6  14; 350; 000  Q i;1 ¼ 0:00268  1147 T g;1  290

3.6. Overall efficiency

Q i;1 ¼ eg rA1 T 4g;1 þ hi A1 ðT g;1  T s;1 Þ ð5Þ

Q H þ Q cooking þ Q i;4 þ Q gw þ Q ga Q i;1 ¼ 0:069  r  0:16  T 4g;1 þ 10  0:16ðT g;1  T s;1 Þ

g¼ _ f  LHV
ð54Þ
m
Q O;1 ¼ ho A1 ðT s;1  290Þ þ erA1 ðT 4s;1  2904 Þ ð6Þ

Q O;1 ¼ 20  0:16ðT s;1  290Þ þ 0:5  r  0:16ðT 4s;1  2904 Þ

Q H : Heat absorbed by the TEG fins, W.

Fig. 12. Stove body.

 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058 1053

Table 2
Characteristics using wood.

Physical parameters Value unit

Net calorific value (LHV) 14,350 [14] kJ/kg
Air–fuel ratio (A/F) 6.63 kg air/kg fuel
_ fÞ
Mass flow rate of fuel ðm 3:52  104 kg/s
Combustor surface area (A1 ) 0.16 m2
Stove surface area before TEG (A2 ) 0.69 m2
Gas Specific heat (Cpg ) 1.147 kJ/kg.K
Combustion efficiency (gc ) 60% [13]
Convection coefficient inside the stove (hi ) 10 [19] W=m2 K
Convection coefficient outside the stove (ho ) 20 [20] W=m2 K
Gas emissivity inside the combustor (eg1 ) 0.069 [19]
Gas emissivity after the combustor (eg2 ) 0.061 [19]
Stove wall emissivity (e) 0.5 [21]
Ambient Temperature (T a ) 290 K
Stefan Boltzmann constant (r) 5:67  108 W=m2 K4

Hence, there are five unknowns namely: (m _ a, m

_ g , T g;1 , T s;1 , and Q i;1 ),
Therefore, Solving Eqs. (1) and (5)–(8), using EES (Engineering
Fig. 13. Dimensions of the TEG hot side fins.
Equation Solver) software, the following values are obtained:

m_ a ¼ 2:33  103 m_ g ¼ 2:68  103 Table 3

Physical parameters of the TEG hot side fins.
T g;1 ¼ 922:9 K T s;1 ¼ 528:5 K
Q i;1 ¼ 1085 W Physical parameters Value Unit
Gas temp. (T g;1 ) 594 K
Number of fins (nf ) 21
Convection coefficient ðhi Þ 10 W=m2 K
Gas emissivity (eg2 ) 0.061 [19]
4.1.1. After combustor zone
Gas emissivity (eg3 ) 0.055 [19]
Stove surface area after TEG fins (A3 ) 0.407 m2
_ f gc LHV  Q i;2 ¼ m
m _ g Cpg ðT g;2  T a Þ ð9Þ Finns surface area (Afin ) 2.02608 m2
Unfinned surface area (Aunfin ) 0.1984 m2
  Fin efficiency (gfin ) 45% [22]
0:000352  0:6  14; 350; 000  Q i;2 ¼ 0:00268  1147 T g;2  290

Q i;2 ¼ eg2 rðA1 þ A2 ÞT 4g;2 þ hi ðA1 þ A2 ÞðT g;2  T s;2 Þ ð10Þ

Q i;2 ¼ ð0:061Þrð0:85ÞT 4g;2 þ ð10Þð0:85ÞðT g;2  T s;2 Þ

Q o;2 ¼ ho ðA1 þ A2 ÞðT s;2  T a Þ þ erðA1 þ A2 ÞðT 4s;2  T 4a Þ ð11Þ

Q o;2 ¼ ð20Þð0:85ÞðT s;2  290Þ þ 0:5rð0:85ÞðT 4s;2  2904 Þ

Hence, there are three unknowns namely: (T g;2 ; T s;2 , and Q i;2 ),
therefore, Solving Eqs. (9)–(11), using EES (Engineering Equation
Solver) software, the following values are obtained:

T g;2 ¼ 594 K T s;2 ¼ 390:4 K

Q i;2 ¼ 2095 W

Fig. 14. Cross section of the TEG model assembly.

4.2. Thermoelectric generator
 
Hot side (see Fig. 13 and Table 3): Q H;conv ¼ 10  0:1984 594  T TEG;fin þ 0:45  10
 
Q H ¼ Q H;rad þ Q H;conv ð12Þ  2:02608 594  T TEG;fin
TEG (see Fig. 14 and Table 4):
Q H;rad ¼ eg rAfins T 4g2 ð13Þ
T TEG;fin  T H
QH ¼ ð15Þ
8 4 Rf ;b þ Ri:p þ Rc:w þ Rms
Q H;rad ¼ 0:061  5:67  10  2:22448  594 ¼ 1083 W
    tb 0:003
Q H;conv ¼ hi Aunfin T g2  T TEG;fin þ gfin hi Afin T g2  T TEG;fin ð14Þ Rf ;b ¼ ¼ ¼ 1:36  103 K=W ð16Þ
kcast:iron Ab 55  0:22
1054 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058

Table 4 Table 5
Physical parameters of the TEG assembly. Physical parameters of the heat sink.

Physical parameters Value Unit Input parameters Value Unit

Interface plate thickness (ti:p ) 1 mm Number of fins (nf ) 13
Ceramic wafer thickness (t c:w ) 1 mm Fin length (L) 0.3 m
Metal solders thickness (t s ) 0.7 mm Fin high (H) 0.035 m
Thermal conductivity of cast iron ðkcast:iron Þ 55 [2] W=m K Width of fin array (W) 0.13 m
Thermal conductivity of aluminum ðkAl Þ 237 [2] W=m K Base thickness ðtb Þ 5 mm
Thermal conductivity of ceramic ðkceramic Þ 25 [2] W=m K Fin thickness (t) 2 mm
Area of the TEG fin base ðAb Þ 0.2 ⁄ 0.2 m Spacing between fins (S) 8 mm
Area of the interface plate ðAi:p Þ 0.2 ⁄ 0.2 m Thermal conductivity of aluminum ðkAl Þ 237 W=m K
Area of the ceramic wafer ðAc:w Þ 0.04 ⁄ 0.04 m Fin emissivity (e) 0.05 [21]
Area of the metal solders ðAs Þ 0.04 ⁄ 0.04 m Fins surface area (Afin ) 0.273 m2
Unfinned surface area (Aunfin ) 0.0288 m2
Fin efficiency (gfin ) 95% [22]
Ambient temperature (T a ) 290 K
ti:p 0:001
Ri:p ¼ ¼ ¼ 9:15  105 K=W ð17Þ Stefan Boltzmann constant (r) 5:67  108 W=m2 K4
kAl Ai:p 237  0:22 Convection coefficient outside the stove (ho ) 20 W=m2 K

t c:w 0:001
Rc:w ¼ ¼ ¼ 2:083  103 K=W ð18Þ
kceramic Ac:w 25  0:042  12 Q c;conv ¼ ho Aunfin ðT HS  T a Þ þ gfin ho Afin ðT Hs  T a Þ ð26Þ
Lms 0:0007
Rms ¼ ¼ ¼ 1:335  104 K=W ð19Þ Q c;conv ¼ ð20Þð0:0288ÞðT HS  290Þ þ ð0:95Þð20Þð0:273ÞðT HS  290Þ
kAl Ams 273  0:042 12
Cold side (heat sink) (see Fig. 15 and Table 5):
T c  T HS
Q c ¼ Q c;rad þ Q c;conv ð20Þ Qc ¼ ð27Þ
Rms þ Rc;w þ RHS;b
 
Q c;rad ¼ nf  1 Q ch T T
   Qc ¼  c HS
þ nf tðL þ 2HÞ þ 2HL þ 2t b ðL þ wÞ re T 4HS  2904
4
ð21Þ 1:538  10 þ 2:083  103 þ RHS;b

 
Q c;rad ¼ ð12ÞQ ch þ ½0:03492 5:67  108 ð0:05Þ T 4HS  2904 0:005
RHS;b ¼ ¼ 5:41  104 K=W ð28Þ
237  0:3  0:13

Electric power (see Table 6):
rðS þ 2HÞL T 4HS  T 4a

Q ch ¼ ð22Þ Ri
ð1  e=eÞ þ ð1=F s-surr Þ Q H ¼ N U pn ðT H  T c Þ þ ISpn T H  I2 ð29Þ
2

rð0:008 þ 2ð0:035ÞÞð0:3Þ T 4HS  2904

Q ch ¼ 10:05 1 6:8
þ ð1=F s-surr Þ Q H ¼ 12 ðT H  T c Þ þ Ið0:054ÞT H  I2
0:05 0:85 2



þ L2 Þ  1
0:5
2H½ð1 Ri
F s-surr ¼ 1  ð23Þ Q c ¼ N U pn ðT H  T c Þ þ ISpn T c þ I2 ð30Þ
 L þ ð1 þ L2 Þ0:5  1
2H 2



    0:3 2 0:5 1 6:8
2 0:035 1 þ 0:008 1 Q c ¼ 12 ðT H  T c Þ þ Ið0:054ÞT c þ I2
0:008 0:85 2
F s-surr ¼1   0:3    0:3 2 0:5 ¼ 0:1238
2 0:035
0:008 0:008
þ 1 þ 0:008 1 I ¼ Spn ðT H  T c Þ=2Ri ð32Þ

I ¼ ð0:054ÞðT H  T c Þ=2  6:8

Pmax ¼ Q H  Q c ¼ ISpn ðT H  T c Þ  I2 Ri ð31Þ

The assembly of the TEG including the hot (fins) and the cold
(heat sink) sides are modeled by Eqs. (12)–(32). These equations
involve: Q H , Q H;rad , Q H;conv , Q c , Q c;rad , Q c;conv , Q ch , T TEG;fin , T H , T HS , T c ,
I, and P max . Therefore, solving these equations using EES (Engineer-
ing Equation Solver) software, the following values are obtained;

Table 6
Physical parameters of the TEG model.

Physical parameters Value Unit

Internal resistance (Ri ) 6.8 X
Number of semiconducting pairs (N) 12
Thermal conductance (U pn ) 1/0.85 W/K
Seebeck coefficient (Spn ) 0.054a V/K
a
Fig. 15. Heat sink of the TEG. Calculated from experiment.
 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058 1055

 
Q cook;conv ¼ ð20Þð0:01176Þ 435  T cook;fin
Q H ¼ 1217 W T TEG;fin ¼ 579:2 K  
þ ð0:55Þð20Þð1:408Þ 435  T cook;fin
Q H;rad ¼ 1083 W T H ¼ 575:2 K
Q H;conv ¼ 133:6 W T HS ¼ 493:9 K
T cook;fin  T cook;p
Q c ¼ 1209 W T c ¼ 496:9 K Q cook ¼ ð41Þ
Rb
Q c;rad ¼ 36 W I ¼ 0:3 X
Q c;conv ¼ 1173 W P max ¼ 7:883 W tb 0:003
Q ch ¼ 2:6 W Rb ¼ ¼ ¼ 1:36  103 ð44Þ
kb Ab 55  0:22

T cook;p  T a T cook;p  290

Q cooking ¼ ¼ ð42Þ
Ac;p ðho þ hr;o Þ 0:22 ð20 þ hr;o Þ
4.2.1. After TEG zone
 
hr;o ¼ re T cook;p þ T a T 2cook;p þ T 2a ð43Þ
_ f gc LHV  Q i;3  Q H ¼ m
m _ g Cpg ðT g;3  T a Þ ð35Þ
 
0:000352  0:6  14; 350; 000  Q i;3  1217 hr;o ¼ 5:67  108  0:7 T cook;p þ 290 T 2cook;p þ 2902
 
¼ 0:00268  1147 T g;3  290 Hence, there are six unknowns namely: Q cook , Q cook;rad , Q cook;conv ,
T cook;fin , T cook;p , and hr;o therefore, Solving Eqs. (38)–(43) using EES
Q i;3 ¼ eg3 rðA1 þ A2 þ A3 ÞT 4g;3 þ hi ðA1 þ A2 þ A3 ÞðT g;3  T s;3 Þ ð36Þ software, the following values are obtained;

Q i;3 ¼ ð0:055Þrð1:257ÞT 4g;2 þ ð10Þð1:257ÞðT g;3  T s;3 Þ Q cooking ¼ 128:5 W T cook;fin ¼ 433 K
Q cook;rad ¼ 112:3 W T cook;p ¼ 432:8 K
Q O;3 ¼ ho ðA1 þ A2 þ A3 ÞðT s;3  T a Þ þ erðA1 þ A2 þ A3 ÞðT 4s;3 Q cook;conv ¼ 16:18 W hr;o ¼ 7:785 W=m2 K

 T 4a Þ ð37Þ

Q O;3 ¼ ð20Þð1:257ÞðT s;3  290Þ þ 0:5rð1:257ÞðT 4s;3  2904 Þ 4.3.1. After cooker zone
Hence, there are three unknowns namely: T g;3 ; T s;3 , and Q i;3 there-
fore, Solving Eqs. (35)–(37). Using EES software, the following val- _ f gc LHV  Q i;4  Q H ¼ m
m _ g Cpg ðT g;4  T a Þ ð45Þ
ues are obtained;
0:000352  0:6  14; 350; 000  Q i;4  1217  128:5
T g;3 ¼ 435 K T s;3 ¼ 336 K  
¼ 0:00268  1147 T g;4  290
Q i;3 ¼ 1370 W

Q i;4 ¼ eg4 rðA1 þ A2 þ A3 þ A4 ÞT 4g;4

þ hi ðA1 þ A2 þ A3 þ A4 ÞðT g;4  T s;4 Þ ð46Þ
4.3. Cooker
Q i;4 ¼ 0:039rð1:53ÞT 4g;4 þ 10ð1:53ÞðT g;4  T s;4 Þ
See Table 7.
Q cook ¼ Q cook;rad þ Q cook;conv ð38Þ Q O;4 ¼ ho ðA1 þ A2 þ A3 þ A4 ÞðT s4  T a Þ
þ erðA1 þ A2 þ A3 þ A4 ÞðT 4s4  T 4a Þ ð47Þ
Q cook;rad ¼ eg rAfins T 4g3 ð39Þ
Q O;4 ¼ 20ð1:53ÞðT s4  290Þ þ 0:5rð1:53ÞðT 4s4  2094 Þ
Q cook;rad ¼ 0:039  5:67  108  1:419  4354 ¼ 112:3 W
Hence, there are three unknowns namely:T g;4 ; T s;4 , and Q i;4
  therefore, Solving Eqs. (45)–(47). Using EES software, the following
Q cook;conv ¼ hi Aunfin T g3  T cook;fin þ gfin hi Afin ðT g2  T cook;fin Þ ð40Þ
values are obtained;

T g;4 ¼ 405 K T s;4 ¼ 327 K

Q i;4 ¼ 1331 W
Table 7
Physical parameters of the cooking model.

Physical parameters Value Unit Table 8

Physical parameters of the water heater.
Gas temp. (T g;2 ) 435 K
Number of fins (nf ) 22 Physical parameters Value Unit
Convection coefficient ðhi Þ 10 W=m2 K Gas temp. (T g;4 ) 405 K
Gas emissivity (eg4 ) 0.039
Water specific heat ðCpw Þ 4187 J=kg K
Stove surface area after TEG fins (A4 ) 0.2795 m 2
Water inlet temperature (T wi ) 290 K
Finns surface area (Afin ) 1.408 m2 Water exit temperature (T wo ) 333 K
Unfinned surface area (Aunfin ) 0.01176 m2 Dew point temperature (T Dp ) 328 K
Fin efficiency (gfin ) 55% [22] Stack surface area (A5 ) 0.235 m
Thermal conductivity of cast iron ðkcast:iron Þ 55 W=m K Gas emissivity (eg ) 0.039
Fin emissivity (e) 0.7 Stack emissivity (estack ) 0.23 [21]
1056 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058

Table 9
4.4. Water heater Characteristics of the three fuels used in the experiment.

Property Fuel
See Table 8
Wood Peat Manure
Q g4—5 ¼ Q gw þ Q ga ð48Þ
Moisture content % 15% 20% 30%
Ash % Trace 4% 11%
_ w Cpw ðT w;i  T w;o Þ
Q gw ¼ m ð49Þ Theoretical air–fuel ratio 5.1 5.39 5.38
Net calorific value (kJ/kg) 14,350 14,500 10,600
_ w ð4187Þð333  290Þ
Q gw ¼ m Initial weight kg 1.6 1.65 1.2
Final weight kg 0.1 0.205 0.233
Time taken to burn it (min) 71 76 66
Q ga ¼ ho A5 ðT stack  T a Þ þ erA5 ðT 4stack  T 4a Þ ð50Þ Ambient temperature °C 17 17 17

Q ga ¼ ð20Þð0:235ÞðT stack  290Þ þ 0:23rð0:235ÞðT 4stack  2904 Þ

The procedure and assumptions of the experiment:
Q ga ¼ eg rA5 T 4g;4 þ hi A5 ðT g;4  T stack Þ ð51Þ
1. The measurements of temperature along different positions of
the stove were taken using type-K thermocouples at different
Q ga ¼ 0:039rð0:235Þ405 þ 10ð0:235Þð405  T stack Þ
4
times in order to compare it with the theoretical data.
_ f gc Q f Þ  Q H  Q cooking  Q i;4 2. The time and the initial and final weight of the fuel were
Q g4—5 ¼ ðm ð52Þ
recorded in order to determine the mass flow rate of fuel.
3. On the TEG assembly, two thermocouples are installed exactly
Q g4—5 ¼ ð3030Þ  1217  128  1331 ¼ 345 W
at the hot and cold side of the TEG model and the measure-
_ w , and T stack
Hence, there are four unknowns namely: Q gw , Q ga , m ments were taken for T H , T c and V oc .
therefore, Solving Eqs. (48)–(52). Using EES software, the following 4. All electrical quantities are measured using digital multi meter
values are obtained; (DMM).
5. All temperature measurements were taken using a type-K ther-
Q ga ¼ 192:4 W T stack ¼ 328 K mocouple digital reader.
Q gw ¼ 152:6 W m_ w ¼ 0:001 kg=s
In general biomass fuel burning goes under three stages (all
three stages can be present at the same time).
4.5. Space heating from stove body
Stage 1: The fuel is heated to evaporate and drive off the
Q heating ¼ Q i;4 þ Q ga þ Q c ð53Þ moisture.
Stage 2: The fuel starts to break down chemically and the vola-
Q heating ¼ 1331 þ 192:4 þ 1209 ¼ 2732:4W tile matter is vaporized. The vapor contains 50–60% of the heat
value of the fuel. This vapor has to be heated to 600 °C to reach
the ignition temperature, hence achieving self-supported flame.
4.6. Overall efficiency
If not, smoke may be generated thus, forming a plume that
envelops the heat exchange surfaces and chimney with
creosote.
Q H þ Q cook þ Q i;4 þ Q gw þ Q ga Stage 3: Once the volatile matter is released, the remaining
g¼ ð54Þ
material (charcoal) burns at temperatures above 800 °C.
m_ f  LHV

1217 þ 128 þ 1331 þ 152 þ 192 Figs. 18–20 show the electric power produced by the TEG mod-
g¼   100% ¼ 59% ules when burning different fuels. The wood produces the highest
3:52  104  14; 350; 000
output power due to its higher temperature but it decays faster
than the peat.
5. Results

The same previous procedure was repeated for peat and man-
ure. The properties of the three fuels are compared in Table 9 as
follows: 1000 Wood Peat Manure

Fig. 16 shows the temperature profile of the gas along the stove. 900
It is obvious from Table 10 that the wood gives the highest gas 800
Temperature K

temperature among other fuels. Although it has lower calorific 700

value than peat but it has a higher burning rate and better perfor- 600
mance. Fig. 17 shows how much energy is recovered in different 500
components of the stove. More than two thirds of the net energy 400
produced in the combustor was recovered in the TEG and in space
300
heating.
200
100
6. Experimental results
0
Combustor TEG Cooker Water heater Exhaust
The experiment is carried out on three types of fuels namely,
wood, peat, and manure with the characteristics shown in Table 9. Fig. 16. Temperature profile along the stove.
 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058 1057

Table 10
Summary of results.

Fuel Q H ðWÞ P max ðWÞ Q heating ðWÞ Q cooking ðWÞ Q gw ðWÞ Q exhaust ðWÞ gov ð%Þ
Wood 1217 7.88 2732 128 152.6 206 59.5
Peat 1013 5.54 2276 110.5 60 176.4 60.4
Manure 616 2.11 1455 77 40 193.8 60

100 Temperature difference Voltage Power

Wood Peat Manure
90 20 12
80 18

Voltage V/Power W
Enegy Recovered %

16 10
70

Temperature °C
14
60 8
12
50
10 6
40 8
30 6 4
20 4
2
10 2
0 0 0
TEG Space heang Cooker Waterheater Exhaust 0 10 20 30 40 50 60 70
Time (min)
Fig. 17. Recovered energy in the components along the stove.
Fig. 20. Output voltage and power plus temperature difference vs time with
manure.
45 Temperature difference voltage power 25
40
350
Voltage V/Power W

35 20 predicted measured
Temperature °C

30 300
15
Temperature °C

25 250
20
10 200
15
10 150
5
5 100
0 0
0 10 20 30 40 50 60 70 50
Time (min)
0
1 TEG 2 cooker 3 Water heater 4
Fig. 18. Output voltage and power plus temperature difference vs time with wood.
Posion

Fig. 21. Comparison of predicted and measured temperature profile along the stove
Temperature difference Voltage Power
45 18 using wood.
40 16
Voltage V/Power W

35 14
Temperature °C

30 12 300
predicted measured
25 10
250
20 8
Temperature °C

15 6 200
10 4
5 2 150
0 0
0 10 20 30 40 50 60 70 100
Time (min)
50
Fig. 19. Output voltage and power plus temperature difference vs time with peat.
0
1 2 3 4

7. Comparison of the system between the measured and posion

predicted data Fig. 22. Comparison of predicted and measured temperature profile along the stove
using peat.
Figs. 21–23 compare the predicted temperature profile with the
measured one. The predicted values are slightly higher than
the measured ones, probably due to the error associated with the constitute. This is in addition to the human error when taking
carbon deposits on the junction of the thermocouple, which the measurements.
1058 Y.S.H. Najjar, M.M. Kseibi / Applied Thermal Engineering 102 (2016) 1045–1058

250 predicted measured References

[1] http://www.worldenergyoutlook.org/resources/energydevelopment/accessto
200
electricity/.
[2] D. Champier, J.p. Bedecarrats, T. Kousksou, M. Rivaletto, F. Strub, P. Pignolet,
Temperature °C

study of TE (Thermoelectric) generator incorporated in a multifunction wood

150
cook stoves, Energy 36 (2011) 1518–1526.
[3] C. Lertsatitthanakorn, Electrical performance analysis and economic evaluation
100 of combined biomass cook stove thermoelectric (BITE) generator, Biores.
Technol. 98 (2007) 1670–1674.
[4] D. Mastbergen, B. Willson, Generating Light from Stoves using a
50 Thermoelectric Generator, Colorado State University, 2007.
[5] P. Raman, N.K. Ram, R. Gupta, Development, design and performance analysis
of a forced draft clean combustion cook stove powered by a thermo electric
0 generator with multi-utility options, Energy 69 (2014) 813–825.
1 2 3 4 [6] http://www.biolitestove.com/
Posion [7] S.M. O’Shaughnessy, M.J. Deasy, J.V. Doyle, A.J. Robinson, Field trial testing of
an electricity producing portable biomass cooking stove in rural Malawi,
Fig. 23. Comarison of predicted and measured temperature profile along the stove Energy Sustain. Develop. 20 (2014) 1–10.
using manure. [8] Y.S.H. Najjar, Combustor liner temperature in the gas turbine using heavy
fuels, J. Inst. Energy LV (425) (1982) 195–200.
[9] Y.S.H. Najjar, Some performance characteristics of the gas turbine combustor
using heavy fuels, J. Inst. Energy LV (425) (1982) 187–194.
[10] Y.S.H. Najjar, E.M. Goodger, Radiation and smoke from the gas turbine
8. Conclusions combustor using heavy fuels, ASME J. Heat Transfer 105 (February) (1983)
82–88.
[11] T.W. Abou-Arab, M.O. Othman, Y.S.H. Najjar, N.T. Ahmad, Combustion and heat
(1) The design and heat transfer performance analysis are car- transfer characteristics for dual-fuel cylindrical water heater model, Fuel 69
ried out for all subcomponents of the stove, over a wide (April) (1990) 485–489.
[12] M.O. Othman, T.W. Abou-Arab, Y.S.H. Najjar, Effect of fuel–air mixture
range of available fuels and operating conditions. heterogeneity on vibration in continuous combustion systems, J. Energy
(2) It is obvious that wood gives the best performance among Environ. 2 (2) (1991) 138–152.
other fuels. It gives the highest gas temperature (920 K) [13] http://www.ruraltech.org/.
[14] J.W. Rose, J.R. Cooper, Technical Data on Fuel, seventh ed., 1977.
and the highest electric power (7.9 W)
[15] S. Rea, S. West, Thermal radiation from finned heat sinks, IEEE Trans. Parts.
(3) The overall efficiency of the stove was about 60% which is Hybrids Pack. 2 (1976) 115–117.
very good compared with the other stoves. [16] N. Ellison, Generalized computations of the gray body shape factor for thermal
radiation from a rectangular U-channel, IEEE Trans. Comp., Hybrids, Manufact.
(4) More than 80% of the energy produced in the combustor is
Technol. 4 (1979) 517–522.
transformed to TEG fins and in space heating (losses from [17] A. Bilitzky, The Effect of Geometry on Heat Transfer by Free Convection From A
the stove walls). Fin Array, MS Thesis, Dept. Of Mechanical Engineering, Ben-Gurion University
(5) Experiments showed that the overall efficiency is obtained of the Negev, Beer Sheva, Israel, 1986.
[18] Y. Shabany, Radiation heat transfer from plate-fin heat sinks, in: Proceedings
with peat due to its high heating value. of 24th Annual IEEE Semiconductor Thermal Measurement and Management
Symposium, SemiTherm, vol. 24, 2008, pp. 133–137.
[19] http://www.cookstove.net/.
[20] N.A. Hudelson, K.M. Bryden, D. Still, Global Modeling and Testing of Rocket
Stove Operating Variations, Department of Mechanical Engineering, Iowa State
Acknowledgment University.
[21] http://www.engineeringtoolbox.com/.
The authors would like to acknowledge the help from Eng. [22] F. Incropera, D.P. Dewitt, T.L. Bergman, A.S. Lavine, Principles of heat and mass
transfer, seventh ed., 2013.
Ahmad M. Abubaker.