Journal of ELECTRONIC MATERIALS

DOI: 10.1007/s11664-013-2545-8 Ó 2013 TMS

Integration of Thermoelectric Generators and Wood Stove to Produce Heat, Hot Water, and Electrical Power A.M. GOUDARZI,1 P. MAZANDARANI,1 R. PANAHI,1 H. BEHSAZ,1 A. REZANIA,2,3 and L.A. ROSENDAHL2 1.—Department of Mechanical Engineering, Babol Noshiravani University of Technology, Babol, Iran. 2.—Department of Energy Technology, Aalborg University, 9220 Aalborg, Denmark. 3.—e-mail: [email protected]

Traditional fire stoves are characterized by low efficiency. In this experimental study, the combustion chamber of the stove is augmented by two devices. An electric fan can increase the air-to-fuel ratio in order to increase the system’s efficiency and decrease air pollution by providing complete combustion of wood. In addition, thermoelectric generators (TEGs) produce power that can be used to satisfy all basic needs. In this study, a water-based cooling system is designed to increase the efficiency of the TEGs and also produce hot water for residential use. Through a range of tests, an average of 7.9 W was achieved by a commercial TEG with substrate area of 56 mm 9 56 mm, which can produce 14.7 W output power at the maximum matched load. The total power generated by the stove is 166 W. Also, in this study a reasonable ratio of fuel to time is described for residential use. The presented prototype is designed to fulfill the basic needs of domestic electricity, hot water, and essential heat for warming the room and cooking. Key words: Thermoelectric generators, wood stove, experimental investigation

INTRODUCTION There are about 2.7 billion people in the world living in rural areas; 1.3 billion of them do not have access to electricity, representing 20% of the world’s total population.1 As it is hard to establish energy supplies in such areas, new energies and practical renewable energies are considered. Traditional or rudimentary stoves, which are widely used as heat generators in developing countries (representing 41% of the total population of the world1), are potential sources to access reliable energy, but they have low efficiency (about 20% to 40%),2 which results in wastage of precious fuel supplies in these areas. Also, these stoves cause air pollution due to incompletely combusted output gases and cause damage to human health. Traditional stoves produce heat by burning fuel, usually solid fuel such as

(Received July 7, 2012; accepted February 13, 2013)

wood or coal. To generate electrical power, one can attach thermoelectric generators (TEGs) to such stoves. We designed a multifunction device that is able to produce a considerable amount of electricity as well as hot water, besides essential heat for cooking and warming a house. Generated electricity can be used to satisfy basic needs such as light, radio, phone, and other electronic devices. This prototype is equipped with a water-based cooling system which also produces hot water. Using TEGs has a lot of advantages compared with other generators, such as:  The stove is designed to reduce the amount of CO2 output, by full combustion of fuel. In fact, there is no pollution while using TEGs, as opposed to traditional stoves.  TEGs are not dependent on sunlight or wind, so unlike solar panels or wind turbines, they can be used at any time of day and any place.  TEGs do not need an auxiliary source of energy. They can use energy exhausted as waste heat.  There is no need to move TEGs. They are silent and work very smoothly.

Goudarzi, Mazandarani, Panahi, Behsaz, Rezania, and Rosendahl

 In this study, to evaluate the thermal efficiency of the designed wood stove, a variety of experiments were carried out. The reported results confirm the thermal behavior of the designed stove. REVIEW OF THERMOELECTRIC GENERATION In 1821, the Estonian physicist Thomas Johann Seebeck discovered that an electromotive force (emf) was produced by a closed loop formed by two metals joined in two places with a temperature difference between the junctions.3 As shown in Fig. 1, a voltage VAB is established between the two junctions of two different metals—A and B—if there is a temperature difference DT between the two junctions. This is the so-called Seebeck effect, which can be characterized by the Seebeck coefficient a, which can be defined as a¼

DV : DT

Employing this effect, a diversity of thermal to electrical convertors have been developed, among which TEGs can be mentioned. The efficiency of such convertors depends not only on the Seebeck coefficient but also on the thermal and electrical conductivities of the convertor components. This can be evaluated by the figure of merit ZT, measured in an identified thermal condition. For technical reasons, to be more efficient, TEGs that are used to convert heat directly to electricity should be composed of several thermoelectric couples connected thermally in parallel and electrically in series. Several combinations of such composite thermoelectric devices have been commercialized as TEG modules. In this study, the designed wood stove is equipped with 21 commercial thermoelectric modules (TEP112656-0.6) that are placed around the walls of the stove at the hottest spots. Each module has 126 couples and size of 56 mm 9 56 mm. The module is

Fig. 1. The Seebeck effect.

bismuth telluride (Bi-Te) based, and can work at heat source temperatures as high as 330°C continuously and up to 400°C intermittently. However, the cold-side temperature of the module cannot go above 200°C. This difference is due to the bonding process, which uses materials with different melting points for each side of the module.4 PROTOTYPE SETUP The implemented wood stove is not catalytic. To burn most of the combustible gases that otherwise would become smoke in traditional stoves, the prototype is designed with baffles and a postcombustion chamber, which route gases through the hottest part of the firebox, mixing them with sufficient air to burn them more completely. When the stove chamber is not initially hot enough to intake fresh air, the stove is equipped with a 1-W fan to aerate the fire. In this prototype the inner part of the stove consists of a two-stage chamber, surrounded by a tight body with two openings to take in fresh air and exhaust the smoke that is a product of the combustion process, which is channelized through the environment of the house or the place where the stove is kept by means of a stovepipe placed behind the stove. Also, there is a third opening to access the main chamber. The latter opening, which is framed by an insulated door, is used to feed the stove and extract the ashes from the main chamber. Hot ashes deposited at the bottom of the main chamber serve to preheat the incoming air. The incoming air stream is canalized to circulate around the combustion chamber through the vacant spaces between the body walls and interior chambers; hence, the circulated air is preheated, which increases the burning efficiency. Since the connection between the chambers and body is slight, heat transferred via air circulation is more significant than heat conducted to the chamber body. There are 21 thermoelectric modules, each attached by its hot side to a 50-mm-long intermediate aluminum spacer attached to the stove body, which prevents damage to the TEGs by exceeding their hot-side temperature limit. To reduce the thermal contact resistance, both sides of each aluminum spacer—which are in contact with the stove body and hot side of the TEGs—are polished with a milling machine to obtain a proper flatness. Each spacer is attached to the stove by two L-shaped bolts. A clamp is designed to tighten each TEG to the spacer surface by two L-shaped bolts which in total generate 430 kg clamping force for optimum power generation and thermal contact of the TEGs (Fig. 2). To prevent heat loss from the stove body, the vacant areas between the spacers are insulated with rockwool material (Fig. 3). The insulated stove body is covered by a thin sheet-metallic case. The modules are divided into seven series, each carrying three modules positioned in a row and

Integration of Thermoelectric Generators and Wood Stove to Produce Heat, Hot Water, and Electrical Power

connected electrically in series. The arrangement of the TEG series around the stove body is shown in Fig. 4. A water-filled aluminum cubic channel with dimensions of 60 mm 9 60 mm 9 25 mm is attached to the cold side of each thermoelectric unit.

Each channel has two separated access pipes located on opposite sides. The access pipes of the grouped modules are connected successively, as are the access pipes of the neighboring groups. Connection of these access pipes is achieved using heatresistant elastic tubes (Fig. 5). Both the water stream passing through the channels and the elastic connections extract heat from the cold side of the TEGs. An overview of the stove assembly is shown in Fig. 6. It should be designed to be heavy to preserve heat in the body frame and optimize the thermal capacity of the stove to act as a heat source (Fig. 7). THEORETICAL DESCRIPTION The voltage measured for a thermoelectric module is defined as5 V ¼ 1np ðTh  Tc Þ

RL ; R þ RL

(1)

where anp is the Seebeck coefficient of the thermocouple, Th and Tc are the temperatures of the hot and cold side of the thermoelectric, and RL and R are the load resistance and internal resistance, respectively. According to Eq. 1, in open-circuit mode when the external resistance tends to infinity, the measured voltage called Voc is equal to Voc ¼ lim 1np ðTh  Tc Þ Fig. 2. Schematic of water cooling system.

RL !1

RL ¼ 1np ðTh  Tc Þ: R þ RL (2)

The electric power consumed in the external resistance is equal to6 P ¼ I 2 RL ¼ I 2 mR ¼

mra2np

ðTh  Tc Þ; (3) ð1 þ mÞ2 l pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi where m ¼ RL =R and mopt ¼ 1 þ ZT ; I is the current intensity; l and r are the leg length and the electrical conductivity of the thermocouple ðr ¼ 1=qÞ. Z¼

a2np a2 r ¼ ; k Knp R

Knp ¼ Kn þ Kp ¼

Fig. 3. Insulation on the stove body.

  An Ap kn þ kp ; ln lp

Fig. 4. The TEG series arrangement.

Goudarzi, Mazandarani, Panahi, Behsaz, Rezania, and Rosendahl

kn ðkp Þ and qn ðqp Þ are the thermal conductivity and electrical resistivity of the n-type (p-type) thermo couple; ln lp and An ðAp Þ are the leg length and cross-sectional P area of the n-type (p-type) thermocouple; and w is the sum of the external thermal resistance. The maximum power generated can be calculated as6 Pmax ¼

Z 4ð1 þ mÞ2

P

w

½2ðTh  Tc Þ2 :

(5)

According to the factory specifications of the thermoelectrics, in the working condition of our setup, the dimensionless figure of merit ðZT Þ is less than 1, so Fig. 5. Water cooling system.

ZT < 1 ! mopt ffi 1: In this study, as we are working with thermoelectrics on a large scale rather than concentrating on describing a physical model for analysis of the thermoelectric behavior, for simplification we neglect thermal and electrical contact effects, which results in  2 /np DT : (6) Pmax ¼ 4R Equating 2 and 6 provides the maximum output power under the matched load condition, forming the basis of the calculations in this study:7 Pmax ¼

Fig. 6. Inner view of the prototype.

2 Voc : 4R

(7)

The thermal energy which the water absorbs by circulating through the cooling system in each stage can be calculated as _ ðTo  Ti Þ; Q_ w ¼ mC

Qw ¼ Q_ w Dt;

where Q_ w is the average rate of heat absorption to _ the water, Qw is the heat absorbed by the water, m is the mass flow rate of water, C is the specific heat capacity of water, Dt is the duration of the stage, and To and Ti are the outlet and inlet temperature of the water, respectively. Also, the energy released from the firewood at each stage can be calculated as Q_ f ¼ Qf =Dt; Fig. 7. The electric circuit of a thermoelectric module.

  ln lp qn þ qp ; (4) An Ap   where Z is the figure of merit; Kn Kp and Rn ðRp Þ are the thermal conductance and the electrical resistance of the n-type (p-type) thermocouple; R ¼ Rn þ Rp ¼

Qf ¼ cm;

where Q_ f is the average rate of heat released from the firewood, Qf is the heat released form the firewood, c is the heat value of the fuel, m is the mass of fuel, and Dt is the duration of the stage. TESTS AND RESULTS All experiments in this work were implemented in an open area under the following conditions:  Average wind velocity: 1.2 m/s  Air pressure: 1 atm

Integration of Thermoelectric Generators and Wood Stove to Produce Heat, Hot Water, and Electrical Power

Fig. 8. Timeline of total Voc for each series of TEGs.

Fig. 9. Timeline of total Pmax for each series of TEGs.

 Input cooling water temperature: 16°C  Pine wood as fuel, with heat value of 18.5MJ/kg8 The open-circuit voltage and temperature of the cold and hot sides of each thermoelectric module were measured periodically using a DMM 390A multimeter and ISOTECH TTI-10 thermometer. These measurement systems enable us to obtain precision of 0.1% for voltage and ±0.012°C for temperature. The results of each experiment were as follows:

 The average rate of heat passed to the water was 3.24 kW.  The average rate of heat released from the firewood was 4.01 kW. The output cooling water flow was 5 L/min, and it took about 15 min to stabilize the thermoelectric output. The test was completed in 60 min, and the residues of wood were weighed. It appeared that only 780 g of wood was burned during this period. Charging 1 kg Firewood per Hour

Preheating with 1 kg Firewood The device was preheated by igniting 1 kg of dry wood. During this experiment, the temperature of the output water was raised to 25.5°C and the temperature at the cold and hot sides of TEGs increased to 28.9°C and 61.3°C, respectively. It can be deduced that:  The maximum power generated by the TEG modules reached 2.025 W.

This experiment was planned to be carried out after the first test and took about 50 min. At the start, the combustion chamber was unloaded and cleaned of ashes and coal, then 1 kg of firewood was replaced. During this test, the temperature of the output water reached 30°C and the temperatures at the cold and hot sides of the TEGs increased to 35.2°C and 178.6°C, respectively. It can be deduced that:

5.74 5.74 1.89 5.74 5.74 4.01 5.139 19.47 30.83 33.63 94.11 3.240508 4.808495 5.453112 9.024639 11.18498 33.71173 0.092054 1.522319 5.004879 13.17866 14.60922 0.166902 1.945264 9.281973 14.31252 18.51605 0.231875 2.870213 12.96927 20.49203 27.92015 Stage Stage Stage Stage Stage Total

1 2 3 4 5

0.181831 2.249223 8.473886 15.07759 22.31627

0.156959 1.838571 7.555997 14.58534 25.95781

0.266652 2.781393 10.09269 14.38506 18.45699

0.20237 2.818152 12.69159 20.26194 27.64226

Series 7 Series 6 Series 5 Series 4 Series 3 Series 2 Series 1

Average Power Generated (W) Table I. Test results

1.298644 16.02513 66.07029 112.2932 155.4188 351.10

Average Rate of Heat Released from Firewood (kW) Average Power Generated During Stage (W)

Average Rate of Heat Absorbed by Water (kW)

Power Consumption by the Pump (W)

Goudarzi, Mazandarani, Panahi, Behsaz, Rezania, and Rosendahl

 The maximum power generated by the TEG modules reached 24.92 W.  The average rate of heat passed to the water was 4.8 kW.  The average rate of heat released from the firewood was 5.139 kW. In this stage, the TEGs made more than six times the energy in 50 min than in the first stage. This indicates that the amount of power generated by the stove and its efficiency strongly depend on whether it is warm or not. These results show the fundamental dependence of the generated power on the amount of wood and the stove temperature. To stabilize the stove, a test was done to estimate the required amount of wood. Charging 4 kg Firewood per Hour This stage (stage 3) started with 1 kg wood, and the output water flow was decreased to 2.5 L/min. After 20 min, another 1 kg of wood was added, but the power generated still fluctuated, so 2 kg of wood was added at 40 min and 70 min later. The results showed that 2 kg of wood added every 30 min was sufficient to stabilize the stove and generate about 70 W of electricity besides hot water at 47.5°C and a temperature of 300°C on top of the prototype for use for cooking or warming the house. Charging 7 kg Firewood per Hour The fourth phase was able to achieve stable generation of 110 W of electricity. The test started with 3 kg of wood in the fire chamber. After 15 min and 30 min, the chamber was refueled with a 1 kg wood package and the output water flow was increased to 5 L/min. So, the average ratio of wood to time for stable electricity generation of 110 W is 7 kg/h. Charging 9 kg Firewood per Hour The final stage was to test the prototype’s ultimate power; therefore, 3 kg of wood was placed in the combustion chamber and then every 15 min 1.5 kg of wood was added. The results showed an average of 7.9 W for each TEG. The maximum power generated by the stove in this step was more than 166 W. Also, the top plate of the stove reached 500°C. Finally, the stove produced hot water at 48.3°C. OVERALL RESULTS The tests consisted of five steps, each of which is interpreted separately. The entire test took more than 5 h. Figures 8 and 9 show the total generated voltage and power during the test, respectively. During the test, 1166.158 kJ of energy was generated by the TEGs and the energy passed to the water was 121,158.4 kJ. To interpret the test results a variety of data are gathered in Table I, which presents the average power generated by each series of TEGs. For each

Integration of Thermoelectric Generators and Wood Stove to Produce Heat, Hot Water, and Electrical Power

stage, the average amount of total power generated is presented. The heat removal by the water and the heat power released from the firewood are also presented in the table. CONCLUSIONS A biomass stove prototype has been completely described, and its prototype setup discussed. A range of tests illustrate that a considerable and stable amount of electricity can be produced by this prototype to satisfy basic needs. The voltage and power produced are shown by graphs, and the total energy is calculated. Also, as aforementioned, the hot water produced by the cooling system is reliable enough for residential use. The water used as the cold source was potable water that was pumped into the cooling system. Although the extra power generated by the TEG depends on the use of a cooling system, comparison with the power that the cooling system consumes is not reasonable; in this study, whose goal is to present a multifunction device, a high mass flow rate of water is an important factor both for cooling the 21 TEGs and for generation of the amount of hot water essential for residential use. So, in this study, the power consumed by the water pump is negligible compared with the total power generated. In rural areas, where these supplements are not accessible, a water tank can be placed at a height of 1 m to provide the water, and gravity will work as the pump to flow water into the cooling system. The hot water produced by the cooling system can then be stored in an insulated tank with glasswool slabs. Due to the high rate of hot water generation, it can be used in a heater radiator system, the output of which can again be used as cold water for the cooling system. Theoretically, the exhaust gases could be flowed around the storage tank in a helical channel to keep the water warm and additionally prevent heat loss, resulting in a further increase in stove efficiency.

All the aims of our laboratory were fulfilled during these tests. Maximum average electricity generation of 9.6 W and 9.4 W was reached by each TEG in the third and fifth series, respectively. In the last stage of the test it reached about 65% of its maximum capability. In similar studies, generally one TE module is used as the generator and the results are generalized under the assumption of the existence of other TEGs. However, it has been observed that even a small displacement of the TEGs can affect the generated power dramatically, so 21 TE modules were used to find the best locations, showing that the results cannot be generalized to other TEGs. However the TEGs that are geometrically polarized give same results. On the other hand, as the number of TEGs increase, the total generated power will also increase, which results in coverage of more residential electricity needs and efficiency increase. To fulfill all the tasks and use the device in developing countries, there are plans to improve the reliability of the device and increase the power generated. Although 166 W is a reasonable amount for basic needs, to mass produce this device the insulation should be improved. REFERENCES 1. International Energy Agency, World Energy Outlook (Paris: IEA Publications, 2011). 2. Cal/EPA Air Resources Board Enforcement Division Compliance Assistance Program, Wood Burning Handbook (Sacramento: California Environmental Protection Agency, 2005). 3. T.J. Seebeck, Abh. K. Akad. Wiss. Berlin 265, 1822 (1821). 4. Thermonamic Company Home Page, Products, www.thermo namic.com/pro_view.asp?id=802. Thermonamic Electronics, Nanchang. Accessed Sept 2012. 5. S. Beeby and N. White, Energy Harvesting for Autonomous Systems (Norwood, MA: Artech House, 2010), pp. 135–157. 6. K. Yazawa and Ali Shakouri, J. Appl. Phys. 111, 024509 (2012). 7. Y. Meydbray, R. Singh, and A. Shakouri, 24th International Conference on Thermoelectrics, ICT 2005 (Piscataway: IEEE, 2005), p. 348. 8. V. Francescato, E. Antonini, and L.Z. Bergomi, Wood Fuels Handbook (Padua: Italian Agriforestry Energy Association, 2008).