Enhanced Efficiency of Thermoelectric Generator by ... - MDPI

energies Article

Enhanced Efficiency of Thermoelectric Generator by Optimizing Mechanical and Electrical Structures Jinlong Chen 1 , Kewen Li 1,2, *, Changwei Liu 1 , Mao Li 3 , Youchang Lv 3 , Lin Jia 1 and Shanshan Jiang 1 1

2 3

*

School of Energy Resources, China University of Geosciences (Beijing), Beijing 100083, China; loongchen@cugb.edu.cn (J.C.); changweiliu@outlook.com (C.L.); jialin@cugb.edu.cn (L.J.); shanshanj@cugb.edu.cn (S.J.) Department of Energy Resources Engineering, Stanford University, Stanford, CA 94305, USA State Key Laboratory of Development and Utilization of Coking-coal Resources, China Pingmei Shenma Group, Pingdingshan 467000, China; zm13721869229@126.com (M.L.); gmc109@126.com (Y.L.) Correspondence: likewen@cugb.edu.cn

Received: 23 July 2017; Accepted: 31 August 2017; Published: 4 September 2017

Abstract: Much attention has been paid to the application of low temperature thermal resources, especially for power generation in recent years. Most of the current commercialized thermal (including geothermal) power-generation technologies convert thermal energy to electric energy indirectly, that is, making mechanical work before producing electricity. Technology using a thermoelectric generator (TEG), however, can directly transform thermal energy into electricity through the Seebeck effect. TEG technology has many advantages such as compactness, quietness, and reliability because there are no moving parts. One of the biggest disadvantages of TEGs is the low efficiency from thermal to electric energy. For this reason, we redesigned and modified our previous 1 KW (at a temperature difference of around 120 ◦ C) TEG system. The output power of the system was improved significantly, about 34.6% greater; the instantaneous efficiency of the TEG system could reach about 6.5%. Laboratory experiments have been conducted to measure the output power at different conditions: different connection modes between TEG modules, different mechanical structures, and different temperature differences between hot and cold sides. The TEG apparatus has been tested and the data have been presented. This kind of TEG power system can be applied in many thermal and geothermal sites with low temperature resources, including oil fields where fossil and geothermal energies are coproduced. Keywords: thermoelectric generator system (TEGs); direct power generation; thermoelectric effect; thermal efficiency

1. Introduction Geothermal energy resources are renewable and exist widely. If carefully managed, geothermal production can be sustained essentially and almost indefinitely [1]. A lot of geothermal resources belong to the low-temperature category, they are being used in combined heat and power plant (CHP). For water with a temperature below 100 ◦ C, binary (Organic Rankine Cycle) power plants are usually installed and used to generate electricity [2,3]. An example is the former USSR binary power plant, 680 KW using 81 ◦ C water at Paratunka on the Kamchatka peninsula [4]. Other methods of waste heat recovery (WHR) that can be applied at low temperature differences with relatively high efficiency, like Texergy cycle [5], trigeneration cycle [6], thermochemical recuperation [7–9], cooling or heating [10,11]. Li et al. [12] discussed possible factors which have led to the low growth rate of geothermal development. The main factors include high exploration risk, lower social acceptance, long payback and construction time and difficulty to assess resources. Li et al. [12] also pointed out the development Energies 2017, 10, 1329; doi:10.3390/en10091329

www.mdpi.com/journal/energies

Energies 2017, 10, 1329

2 of 15

and utilization of thermoelectric generator (TEG) technology may be one of the possible solutions and a direction to speed up geothermal growth. Liu et al. [13] built a power generator using TEG modules, which indicated that the cost of the TEG system developed was lower than those of photovoltaic (PV) and wind power systems in terms of equivalent energy generated. From the beginning of the 1990s a renewal of interest in thermoelectric (TE) appeared, due to environmental concern with refrigerant gas and greenhouse effects, as well as the need to develop alternative energy sources [14]. One of the big disadvantages of TEGs is the low efficiency from thermal to electric energy because of many factors affecting the output power of thermoelectric generators. In the last few decades, immense efforts have been made to study thermoelectric generator applications experimentally. Zheng et al. [15] reviewed the thermoelectric technology, demonstrated its potential to improve energy efficiency and pointed to a possible direction of alleviating energy demand. Due to the advantages of no moving parts, long lifespan and quiet operation, attempts to use thermoelectric modules have been made in many areas, such as the automobile, aerospace, industry and domestic sectors. Demir and Dincer [16] investigated a water-electric cogeneration system integrated with a TEG, the system has 21.8 MW of generating electricity capacity. Niu et al. [17] developed a low-cost, simple configuration TEG unit with commercially available Bi2 Te3 based thermoelectric modules for an anticipated maximum power generation of about 150–200 W level. Chen et al. [18] researched the characteristics of a multi-element thermoelectric generator with the irreversibility of finite-rate heat transfer, Joule heat inside the thermoelectric device, and the heat leak through the thermoelectric couple leg. Crane and Jackson [19] studied optimizing TE waste heat recovery by integrating efficient cross flow heat exchangers with thermoelectric modules for conversion of waste heat to electricity. Hsu et al. [20] made efforts in simulating and testing the performance of Bi2 Te3 -based TEG modules, 12.41 W of maximum power output at an average temperature difference of 30 K for 24 TEG modules. Rodríguez et al. [21] developed a TEG model which takes into account the properties of the thermoelectric materials as a function of temperature, studied the behavior of the thermoelectric generator and made improvements to the previously published model. Bélanger and Gosselin [22] presented a model and optimized the internal structure of a thermoelectric generator sandwiched in a cross flow heat exchanger. Ming et al. [23] carried out a numerical simulation to verify the validity of the optimum segmented thermoelectric generator (STEG) model in the design boundary condition. The results indicated that unsteady heat flux not only affects the output of the model, but also deteriorates the work environment. Esarte [24] developed a theoretical expression for the temperature through the TEG module and checked with the experimental results. The theoretical results meet well with the experimental values for low flow rates but not for high flow rates. Gou et al. [25] studied the influence of heat transfer irreversibility on thermoelectric generation performance. Expanding the heat sink surface area and enhancing cold-side heat transfer capacity in a proper range can enhance performance of the TEG system. Casano and Piva [26] reported an experimental investigation of the performance of a power generation device in which they used multiple Peltier modules in the Seebeck mode, and analyzed the thermoelectric generator based on the experimental data for the ‘open’ and ‘closed’ circuit voltage, electric power output and conversion efficiency as a function of temperature. Chen et al. [27] presented a three-dimensional (3D) numerical solution to the fluid-structure coupled problem by implementing a FLUENT compatible thermoelectric model. Yu and Zhao [28] presented a detailed numerical model of thermoelectric generator with the parallel-plate heat exchanger, focused on analyzing the fluid temperature changes along the fluid passage and the temperature difference cross the thermoelectric modules (TE). Meng et al. [29] established a complete numerical model of commercial thermoelectric generator with finned heat exchangers, considered the heat transfer irreversibilities between the device and its external reservoirs, analyzed the effects of external irreversibilities on the performance of the thermoelectric generators compared with the exo-reversible model. Many researches have investigated the numerical modeling of TEG systems. Gou et al. [30] established a theoretical dynamic model of a thermoelectric generator for waste heat recovery and

Energies 2017, 10, 1329

3 of 15

Gou et al. [30] established a theoretical dynamic model of a thermoelectric generator for waste heat recovery and discussed the influence of the flow rates of hot and cold fluids, which indicated that Energies 2017, 1329 3 of 15 heat enhancing heat10,dissipation on the cold side takes greater importance compared with enhancing transfer. In this paper, we redesigned and modified our previous 1 KWindicated (at a temperature difference discussed the influence of the flow rates of hot and cold fluids, which that enhancing heat of around 120 °C)on TEG output power of the system was improved significantly, dissipation thesystem. cold sideThe takes greater importance compared with enhancing heat transfer.about 34.6% paper, we redesigned and of modified our previous KW (atreach a temperature difference of greater; In thethisinstantaneous efficiency the TEG system 1could about 6.5%. Laboratory ◦ C) TEG system. The output power of the system was improved significantly, about 34.6% around 120 experiments have been conducted to measure the output power at different conditions: different greater; the instantaneous of the TEGdifferent system could reach aboutstructures, 6.5%. Laboratory experiments connection modes betweenefficiency TEG modules, mechanical different fluid flow have been conducted to measure the output power at different conditions: different connection modes direction, and different temperature differences between hot and cold sides. The TEG apparatus has TEG modules, different mechanical structures, different fluid flow direction, and different beenbetween tested and the data will be presented. temperature differences between hot and cold sides. The TEG apparatus has been tested and the data will be presented.

2. Experimental

2.AExperimental schematic diagram of the experiment is shown in Figure 1. The temperature on the hot side was keptAatschematic a given constant value (40, 50, 60,is70, 80, 90 or 100 1. °C) bytemperature a thermostat diagram of the experiment shown in Figure The on water the hotbath, side and ◦ C) by a thermostat water bath, and was kept a given constanttemperature value (40, 50,was 60, 70, 80,at90about or 10020 the cold side at with a constant kept °C, which was controlled by the tap theThe coldtemperature side with a constant temperature at about 20 ◦ C,ofwhich controlled by and the five water. of hot and cold sideswas waskept within a margin errorwas of three percent tap water. The temperature of hottemperatures and cold sides of was within a margin errorwere of three percent and percent respectively. The average the hot and coldofsides measured by two five percent respectively. The average temperatures of the hot and cold sides were measured by two micro-thermocouples. The values of voltage, current and power were collected by the multi-meter micro-thermocouples. The values of voltage, current and power were collected by the multi-meter devices. Incremental conductivity method was used to achieve the maximum power tracking as the devices. Incremental conductivity method was used to achieve the maximum power tracking as load resistance was the optimum value [31,32]. In this study, the temperature of the hot side is the load resistance was the optimum value [31,32]. In this study, the temperature of the hot side is represented by by Th, Tand the cold side is Tc, in addition, the temperature difference between hot and represented h , and the cold side is Tc , in addition, the temperature difference between hot and coldcold sides is referred to asas ∆T TcT).c ). sides is referred to ∆T(∆T (∆T==TTh h−−

Figure thermoelectricgenerator. generator. Figure1.1.Schematic Schematic of of the thermoelectric

3. Characteristics Internal Resistance 3. Characteristics of of Internal Resistance internal resistance of of the module affects the current that goesthat through finally it, TheThe internal resistance thethermoelectric thermoelectric module affects the current goesit,through determining the output power. Therefore, it is crucial to the internal resistance characteristics. finally determining the output power. Therefore, it study is crucial to study the internal resistance ◦ C with an interval of 10 ◦ C, The ∆T of single thermoelectric modules was increased from 50 to 90 characteristics. the experimental results are shown in Figure 2. The ∆T of single thermoelectric modules was increased from 50 to 90 °C with an interval of 10 °C, the experimental results are shown in Figure 2.

Energies 2017, 10, 1329 Energies 2017, 2017, 10, 10, 1329 1329 Energies

4 of 15

of 15 15 44 of

1.5 1.5

∆T T=9 =900℃ ℃ ∆ ∆T T=8 =800℃ ℃ ∆ ∆ T =7 0 ℃ ∆ T =7 0 ℃ ∆T T=6 =600℃ ℃ ∆ ∆T T=5 =500℃ ℃ ∆

Output OutputPower, Power,W W

1.2 1.2

0.9 0.9

0.6 0.6

0.3 0.3

0.0 0.0

00

11

22

33

44 55 Load, Ω Ω Load,

66

77

88

99

Figure 2.2. The relationship between load and outputpower powerfor forsingle singlemodule module at different ∆T. Figure The relationshipbetween betweenload load and and output output atat different ∆T. Figure 2. The relationship power for single module different ∆T.

AsAs shown ininFigure Figure 2,2,aaasmall-scale small-scale increase of output outputpower powerappears appearsatat atthe theinitial initial development As shown of output power appears the initial development shownin Figure2, small-scale increase increase of development period. With the increaseinin inload, load, the output power increases firstdecreases then decreases decreases gradually. The period. load, output power increases first then gradually. The period.With Withthe the increase increase thethe output power increases first then gradually. The output output power comes to the maximum when the load is equal to the internal resistance, these results output to the maximum the loadtoisthe equal to the internalthese resistance, these results powerpower comescomes to the maximum when thewhen load is equal internal resistance, results conform to conform to the Closed Circuit Ohm’s Law. The dash line describes the optimum resistance that has conform to the Closed Circuit Law. dash line the optimum resistance that has the Closed Circuit Ohm’s Law.Ohm’s The dash lineThe describes the describes optimum resistance that has an increasing ◦ an increasing tendency at different ∆T, varying from 50 to 90 °C. It seems to be a phenomenon that antendency increasing at different 50seems to 90 to °C.beIt aseems to be a phenomenon that at tendency different ∆T, varying ∆T, fromvarying 50 to 90from C. It phenomenon that increasing increasing temperature difference leads to the internal resistance of thermoelectric module increasing temperature difference leads to the internal resistance of thermoelectric module increasing due to the increasing temperature difference leads to the internal resistance of thermoelectric module increasing due to the the characteristics characteristics of thermoelectric thermoelectric materials intemperature specific temperature temperature range [33]. [33]. characteristics of thermoelectric materials in a specificin range [33]. range due to of materials aa specific In addition, the resistances of multiple thermoelectric modules were tested at different ∆T, In addition, the resistances of multiple thermoelectric modules were tested at ∆T, In addition, the resistances of multiple thermoelectric modules were tested atdifferent different ∆T, ◦ increasingfrom from20 20 to C, and the studies on aon number of thermoelectric modules in seriesin increasing from 20 to 80 80 °C, °C, andthen then the studies on number of thermoelectric thermoelectric modules inwere series increasing 80 and then the studies aa number of modules series conducted. The experimental results are presented in Figures 3 and 4. were conducted. The experimental results are presented in Figures 3 and 4. were conducted. The experimental results are presented in Figures 3 and 4. 120 120

80 80

∆T=60℃ ∆T=60℃ ∆T=40℃ ∆T=40℃

Internal InternalResistance, Resistance,ΩΩ

100 100

∆T=80℃ ∆T=80℃ ∆T=70℃ ∆T=70℃

∆T=30℃ ∆T=30℃ 60 60 40 40 20 20 00

00

55

10 10

15 15

20 25 30 20 25 30 The Quantity Quantity of of Modules Modules The

35 35

40 40

45 45

Figure 3.3.Relationship Relationship between total internal resistanceand andquantity quantityofof ofmodules modules in series. Figure Relationshipbetween betweentotal total internal internal resistance inin series. Figure 3. resistance and quantity modules series.

Energies 2017, 10, 1329 Energies 2017, 10, 1329

5 of 15

5 of 15

120 n=40

100

n=20

Internal Resistance, Ω

n=8 80

n=6 n=4

60

n=3 n=2 n=1

40 20 0 10

20

30

40

50 60 70 Temperature Difference, ℃

80

90

100

Figure Relationshipbetween betweeninternal internal resistance resistance and modules (n (n in in thethe Figure 4.4.Relationship and ∆T ∆Tfor fordifferent differentquantity quantityofof modules diagram is the number of thermoelectric modules in series). diagram is the number of thermoelectric modules in series).

Figure3 3demonstrates demonstratesaalinear linearrelationship relationship between and thethe quantity Figure betweenthe thetotal totalinternal internalresistance resistance and quantity moduleswhen when∆T ∆Tisisconstant. constant. Figure Figure 44 reveals the forfor a a of of modules the fluctuation fluctuationrange rangeofofinternal internalresistance resistance ◦ C. Obviously, for 4 modules, the different number of modules, when ∆T is increasing from 20 to 80 different number of modules, when ∆T is increasing from 20 to 80 °C. Obviously, for 4 modules, the ◦ C; maximum fluctuationrange rangeof oftheir their internal internal resistance ofof 4040 Min maximum fluctuation resistance isisabout about5.6% 5.6%(R(R Min==8.9 8.9ΩΩatata a∆T∆T °C; ◦ R = 11.6 Ω at the ∆T of 80 C); for 20 modules, it is around 10.6%, and for 40 modules, 15.3%. Max RMax = 11.6 Ω at the ∆T of 80 °C); for 20 modules, it is around 10.6%, and for 40 modules, 15.3%. The resultsillustrate illustrate that that the ofof total resistance for multiple modules increases as the as The results the fluctuation fluctuationrange range total resistance for multiple modules increases increases. The resistivity ofof the thermoelectric material is changed in in a specific thenumber numberofofmodules modules increases. The resistivity the thermoelectric material is changed a specific temperature range [33], and the increasing quantity of modules may amplify the internal resistance temperature range [33], and the increasing quantity of modules may amplify the internal resistance fluctuation. Therefore, stabilizing the internal resistance may be a key issue for thermoelectric power fluctuation. Therefore, stabilizing the internal resistance may be a key issue for thermoelectric power generation development. generation development. 4. Factors Affecting Output Power

4. Factors Affecting Output Power

Today, many studies regarding thermoelectricity are focused on material improvement. Today, many regarding thermoelectricity focused onThe material improvement. Less Less attention hasstudies been paid to the device and systemare optimization. heat exchanger system attention has important been paid part to the and measurement system optimization. The heat exchanger system is the most is the most indevice optimized and the TEG system, therefore, experimental important part in optimized measurement and the TEG system, therefore, experimental study was study was focused on the structure, reinforcement, insulation measures, fluid flow direction, and flow focused reinforcement, measures, fluid direction, and flow rate of on thethe heatstructure, exchanger, by providing ainsulation relevant experimental basisflow for the following study onrate the of module. theTEG heat exchanger, by providing a relevant experimental basis for the following study on the TEG

module.

4.1. Effect of Heat Exchanger Structure

4.1. Effect Heatexchanger Exchangersurface Structure Theofheat has a direct relationship with the contact thermal resistance between heat exchanger and thermoelectric module. In addition, the fluid channel’s shape of the heat exchanger The heat exchanger surface has a direct relationship with the contact thermal resistance between also affects the heat transfer efficiency. Therefore, three kinds of plate heat exchanger models with the heat exchanger and thermoelectric module. In addition, the fluid channel’s shape of the heat same material but different structures were selected as the research object whose properties can be exchanger also affects the heat transfer efficiency. Therefore, three kinds of plate heat exchanger seen from the data in Table 1. The schematic diagram and the structures of three models are shown in models with the same material but different structures were selected as the research object whose Figures 5 and 6.

properties can be seen from the data in Table 1. The schematic diagram and the structures of three models are shown in Figures 5 and 6.

Energies 2017, 2017, 10, 10, 1329 1329 Energies Energies 2017, 10, 1329

6 of 15 6 of 15

Figure 5. Different model structure: (a) Structure diagram of Model #1; (b) Structure diagram of Model Figure 5. Different Different model structure: Structure diagram of Model #1;Structure (b) Structure diagram of Figure 5. model structure: (a)(a) Structure diagram of Model #1; (b) diagram of Model #2; (c) Structure diagram of Model #3. Model #2; (c) Structure diagram of Model #3. #2; (c) Structure diagram of Model #3.

Figure 6. (a) Contact form between module and heat exchanger of Model #1 and #2; (b) Contact form Figure 6.module (a) Contact between module and#3. heat exchanger of Model #1 and #2; (b) Contact form between and form heat exchanger of exchanger of Model Model #3. between module and heat exchanger of Model #3.

Model Model Model Model Model #1 Model #1 Model #1 Model #2 Model #2 Model #2 Model #3#3 Model #3

Fluid Channel Fluid Channel Shapes Fluid Channel Shapes Shapes Through-hole Through-hole Through-hole N-type boxes N-type boxes N-type boxes Rough Roughsurface surface boxes Rough surface boxes boxes

Table 1. 1. The structure structure comparison comparison of of Models Models #1, #1, #2 #2 and and #3. #3. Table Table 1. The structure comparison of Models #1, #2 and #3. Quantity Quantity ofQuantity Module of Module of Module 40 40 40 40 40 40 40 40 40

Dimensions Fluid Channel Dimensions Fluid Channel (mm) Cross-Sectional Area (cm2) Dimensions Fluid Channel (mm) Cross-Sectional Area (cm2 ) (mm) Cross-Sectional Area (cm2) 1000 × 100 × 30 3.14 1000 × 100 × 30 3.14 1000 × 100 × 30 3.14 1000 × 100 × 50 36 1000 × 100 × 50 36 1000 × 100 × 50 36 1000 1818 11.28 1000 ××100 100×× 11.28 1000 × 100 × 18 11.28 1 Volume: 1 Volume:the offluid fluidflow flow channel. thevolume volume of channel. 1

Wall Thickness Wall Thickness Wall (mm) Thickness (mm) (mm) 5 5 5 5 5 5 33 3

Volume 1 Volume 1 1 (mL) Volume (mL) (mL) 314 314 314 3600 3600 3600 1128 1128 1128

Volume: the volume of fluid flow channel.

The output output power power of of Models Models #1, #1, #2 #2 and tested at at different different ∆T ∆T from from 40 40 to to 80 80 ◦°C, and the the The and #3 #3 is is tested C, and The output power of Models #1, #2 and #3 is tested at different ∆T from 40 to 80 °C, and the results are shown in Figure 7. results are shown in Figure 7. results are shown in Figure 7.

Energies 2017, 10, 1329 Energies 2017, 10, 1329

7 of 15 7 of 15

Output Power, W

40

Model #3

35

Model #2

30

Model #1

y = 0.5159x - 11.987 R² = 0.989

y = 0.5442x - 10.13 R² = 0.9874

25 20 15 y = 0.5187x - 13.811 R² = 0.9989

10 5 0 30

40

50

60

70

80

90

100

Temperature Difference, ℃

Figure 7. Relationship between ∆T ∆T and maximum output power in different models.

Figure 7 indicates that the maximum output power of Models #1, #2 and #3 is 7.1, 8.6 and 10.45 W Figure 7 indicates that the maximum output power of Models #1, #2 and #3 is 7.1, 8.6 and 10.45 W respectively for 40 thermoelectric modules at a ∆T of 40 ◦°C. When the ∆T increases to 80 ◦°C, the respectively for 40 thermoelectric modules at a ∆T of 40 C. When the ∆T increases to 80 C, the maximum output power of Models #1, #2 and #3 goes up to 27.7, 30.3, and 32.5 W respectively. These maximum output power of Models #1, #2 and #3 goes up to 27.7, 30.3, and 32.5 W respectively. results suggest that the overall performance of Model #3 is more stable than that of Models #1 and These results suggest that the overall performance of Model #3 is more stable than that of Models #1 #2. All the improvements of the heat exchanger are embodied in the structure design, and reflected and #2. All the improvements of the heat exchanger are embodied in the structure design, and reflected in the level of power change of a heat exchanger model, which was composed of 40 thermoelectric in the level of power change of a heat exchanger model, which was composed of 40 thermoelectric modules only. The advantage of Model #3 will be enlarged greatly when a series of heat exchangers modules only. The advantage of Model #3 will be enlarged greatly when a series of heat exchangers are assembled for a TEG system. are assembled for a TEG system. As shown in Figure 5a, the fluid channels of Model #1 are connected in series; the fluid flow As shown in Figure 5a, the fluid channels of Model #1 are connected in series; the fluid flow direction was selected co-current (the heat fluid flow direction is the same as the cold fluid flow direction was selected co-current (the heat fluid flow direction is the same as the cold fluid flow direction). The structure of Model #1 is relatively simple and the thicker wall limits the heat conduction, direction). The structure of Model #1 is relatively simple and the thicker wall limits the heat conduction, which reduces the thermoelectric conversion efficiency of the whole model. Thermoelectric modules which reduces the thermoelectric conversion efficiency of the whole model. Thermoelectric modules on the surfaces of Models #1 and #2 are relatively loose, which can be seen from Figure 6a. on the surfaces of Models #1 and #2 are relatively loose, which can be seen from Figure 6a. According to Figure 5b, the structure of Model #2 is comparatively reasonable, the larger volume According to Figure 5b, the structure of Model #2 is comparatively reasonable, the larger volume of fluid channel cross-sectional area and the thinner wall thickness contribute to the heat conduction, of fluid channel cross-sectional area and the thinner wall thickness contribute to the heat conduction, which may improve the overall thermoelectric conversion efficiency to a certain extent. Nevertheless, which may improve the overall thermoelectric conversion efficiency to a certain extent. Nevertheless, the surface of the fluid channel is smooth, which may not help to increase heat transfer efficiency. the surface of the fluid channel is smooth, which may not help to increase heat transfer efficiency. Compared with Model #1, the problems of poor compactness between thermoelectric modules and Compared with Model #1, the problems of poor compactness between thermoelectric modules and heat exchanger in Model #2 are still not improved. heat exchanger in Model #2 are still not improved. Based on Model #1 and Model #2, Model #3 employs embedded methods, and its structure Based on Model #1 and Model #2, Model #3 employs embedded methods, and its structure diagram is shown in Figure 5c. The thermoelectric modules were embedded in the bulge which was diagram is shown in Figure 5c. The thermoelectric modules were embedded in the bulge which was installed in the middle of the heat exchanger plate, as shown in Figure 6b, and the space between the installed in the middle of the heat exchanger plate, as shown in Figure 6b, and the space between the two bulges placed in two rows of the thermoelectric modules. The structure of Model #3 ensures the two bulges placed in two rows of the thermoelectric modules. The structure of Model #3 ensures the compactness between thermoelectric modules and heat exchanger when the pressure is applied to compactness between thermoelectric modules and heat exchanger when the pressure is applied to the the heat exchanger, the thinner wall can increase heat conduction, the rough treatment inside the heat exchanger, the thinner wall can increase heat conduction, the rough treatment inside the fluid fluid channel makes the contact area of the fluid and the fluid channel increase greatly, all these channel makes the contact area of the fluid and the fluid channel increase greatly, all these measures measures contribute to the increasing conversion efficiency of TEG. contribute to the increasing conversion efficiency of TEG. 4.2. Effect 4.2. Effect of of Heat Heat Exchanger Exchanger Reinforcement Reinforcement The TEG TEG system system requires requires aa high high degree degree of of compactness; compactness; besides besides conventional conventional fixed fixed methods methods The which just reinforced pressure at either side of heat exchanger, the reinforcement measures also which just reinforced pressure at either side of heat exchanger, the reinforcement measures also added added pressure at thepart. central maythe improve the stress distribution uniformity and the pressure at the central Thispart. may This improve stress distribution uniformity and the compactness compactness between thermoelectric modules and heat exchanger system. The value of ∆T between thermoelectric modules and heat exchanger system. The value of ∆T was controlled was and controlled and ranged from 40 to 80 °C, the results of the maximum output power before and after the application of reinforcement measures at the same ∆T for Model#1 is shown in Figure 8.

Energies 2017, 10, 1329

8 of 15

ranged from 40 to 80 ◦ C, the results of the maximum output power before and after the application of Energies 2017, 10, 1329 8 of 15 reinforcement measures at the same ∆T for Model#1 is shown in Figure 8. 30 y = 0.3914x - 6.6968 R² = 0.9869

Output Power, Power, W W Output

25 20 15

y = 0.3813x - 3.1491 R² = 0.9877

10

With reinforcement measures

5

Without reinforcement measures 0

30

40

50

60

70

80

90


Figure 8. The effect of reinforcement measures power. measures on on output output power.

Figure 88 indicates indicates that that the the maximum maximum output output power power is 12.5 W W without without reinforcement reinforcement measures, Figure is 12.5 measures, and 16.4 W with reinforcement measures when the ∆T is about 50 °C. The maximum output power ◦ and 16.4 W with reinforcement measures when the ∆T is about 50 C. The maximum output power is is 23.7 without reinforcement measures, and 26.5 with reinforcement measurements when 23.7 WW without reinforcement measures, and 26.5 WW with reinforcement measurements when ∆T∆T is is about °C. According thedata dataabove, above,the thecompactness compactnessbetween betweenthe thethermoelectric thermoelectric modules modules in in ◦ C. about 8080 According totothe Model #1 was improved and the phenomenon of air gap was alleviated owing to the thermal Model #1 was improved and the phenomenon of air gap was alleviated owing to the thermal resistance resistance decreased taking reinforcement the thermoelectric decreased after takingafter reinforcement measures. measures. Therefore, Therefore, the thermoelectric conversionconversion efficiency efficiency and the output power of Model #1 were increased. and the output power of Model #1 were increased. 4.3. Effect Effect of of Heat Heat Exchanger Exchanger Insulation Insulation Measurements Measurements 4.3. A low low temperature temperature environment environment leads a large large amount amount of of heat heat loss loss and and aa low low heat heat energy energy A leads to to a utilization rate. carrier toto study thethe influence of utilization rate. Model Model#2, #2,as asshown shownininFigure Figure5b, 5b,was wasselected selectedasasthe the carrier study influence heat-insulation measurements on output power, the insulation material was filled in the space of heat-insulation measurements on output power, the insulation material was filled in the space between the theheat heatexchanger exchanger and thermoelectric modules thoroughly, then tested the output between and thethe thermoelectric modules thoroughly, then tested the output power power respectively with insulation material or not. Heat insulation measurements can improve the respectively with insulation material or not. Heat insulation measurements can improve the output output power about 5% according to the data listed in Figure 9. power about 5% according to the data listed in Figure 9. 35

Output Power, Power, W W Output

y = 0.5478x - 14.268 R² = 0.9981

With insulation

30

Without insulation

25 20 15

y = 0.496x - 8.9829 R² = 0.9928

10 5 0

30

40

50

60

70

80


Figure 9. power. 9. The effect of heat-insulation heat-insulation on on output output power.

90

Energies 2017, 10, 1329

9 of 15

Energies 2017, 10, 1329

9 of 15

4.4. 4.4. Effect Effect of of Thermoelectric Thermoelectric Module Module Connection Connection Methods Methods Thermoelectric Thermoelectric modules modules are are connected connected in in series series or or parallel. parallel. In In practical practical applications, applications, when when the the same number of thermoelectric modules are connected, the more the thermoelectric modules same number of thermoelectric modules are connected, the more the thermoelectric modules are are in in series, the higher voltage is generated in the closed circuit, which reduces the voltage loss. series, the higher voltage is generated in the closed circuit, which reduces the voltage loss. Model Model #1 #1 was was chosen chosen as as carrier carrier to to study study the the effect effect of of the the connection connection method method between between the the ◦C thermoelectric modules on the output power. In the experiments, the ∆T was maintained thermoelectric modules on the output power. In the experiments, the ∆T was maintained at at 80 80 °C unchanged, the connection connectionmethod methodwas wasselected selected Table 2, and output power obtained. unchanged, the asas Table 2, and thethe output power waswas obtained. The The results are demonstrated in Figure 10. results are demonstrated in Figure 10. Table 2. Different connection methods among 40 modules.

Table 2. Different connection methods among 40 modules. Mode Mode Mode 11 Mode 22 Mode Mode 33 Mode 44 Mode 55

ConnectionMethods Methods Connection 55modules in series as one group, modules in series as one group,88groups groupsas asparallel parallel 88modules modulesin inseries seriesas asone onegroup, group,55groups groupsas asparallel parallel 10 10modules modulesin inseries seriesas asone onegroup, group,44groups groups as as parallel parallel 20modules modulesin inseries seriesas asone onegroup, group,22groups groups as as parallel parallel 20 40modules modulesin inseries series 40

25

Ⅰ

Output Power, W

20

15 Mode 1 Mode 2 Mode 3 Mode 4 Mode 5

10

∆R4 5 0

20

40

60 Load, Ω

80

100

120

140

Figure 10. 10. Relationship Relationship between between output output power power and and load load for for 40 40 modules modules with with different different connection connection Figure methods (∆T (∆T is is 80 80 ◦°C). methods C).

The power ofof Modes 1, 1, 2, 2, 3, The experimental experimentaldata dataare arebriefly brieflysummarized summarizedasasfollows: follows:the thepeak peakoutput output power Modes 43,and 5 is 22.57, 22.1, 22.46, 22.95 and 22.79 W respectively, and the corresponding optimum resistances 4 and 5 is 22.57, 22.1, 22.46, 22.95 and 22.79 W respectively, and the corresponding optimum are 1.18, 4.38, 88.29 Ω.and 88.29 Ω. resistances are7.59, 1.18,26.56 4.38,and 7.59, 26.56 In In order order to to study study how how the the corresponding corresponding load load changes changes when when the the peak peak output output power power is is reduced reduced about about 10%, 10%, we we intercept intercept the the corresponding corresponding load load for for the the 20 20 W W shown shown as as straight straight line line II (Figure (Figure 10), 10), and and the the results results are are listed listed in in Table Table3.3. Table Table 3. 3. Different Different ∆R ∆Rii for for different different connection connection methods. methods.

Resistance Mode11 Mode 2 2Mode 3 Mode 4 Mode 5 Resistance Mode Mode Mode 3 Mode 4 Mode 5 Optimum resistance value (Ω) 1.18 4.38 7.59 26.56 88.29 Optimum resistance value (Ω) 1.18 4.38 7.59 26.56 88.29 ∆Ri 1 3.06 5.03 7.66 36.53 50.8 3.06 5.03 7.66 36.53 50.8 ∆Ri 1 1 ∆Ri represents the variation range in the value of the resistance, when the power range changes 10%. 1 ∆Ri represents the variation range in the value of the resistance, when the power range changes 10%.

It can can be be seen seen from from Table Table 33that that ∆R ΔR11 is is 3.06 3.06 Ω Ω as as 40 40 thermoelectric thermoelectric modules modules connected connected in 1, It in Mode Mode 1, while ∆R ΔR55 is is 50.8 50.8 Ω Ω as as 40 40 thermoelectric thermoelectric modules modules connected connected in in Mode Mode 5, 5, and and ∆R ∆Rii also also rise rise with with the the while increasing number number of of thermoelectric thermoelectric module module in in series. series. These These experimental experimental results results indicated indicated that that the the increasing connection modes between the thermoelectric modules may not affect the peak power of TEG when the quantity is the same.

Energies 2017, 10, 1329

10 of 15

Energies 2017, 10,modes 1329 between the thermoelectric modules may not affect the peak power of TEG when 10 of 15 connection Energies 2017, 10, the quantity is1329 the same.

10 of 15

4.5. Effect of Fluid Flow Direction 4.5. 4.5. Effect Effect of of Fluid Fluid Flow Flow Direction Direction Figure 11a shows the case in which the heat fluid flow direction is the same as the cold fluid Figure shows the which heat flow is as cold Figure 11a 11a shows the case case in in and which the heat fluid fluid flow direction direction is the the same same11b). as the the cold fluid fluid flow direction, that is, co-current, thethe opposite is counter current (Figure Owing to the flow direction, that is, co-current, and the opposite is counter current (Figure 11b). Owing to the flow direction, that is, co-current, and the opposite is counter current (Figure 11b). Owing to the thinner wall of Model #2 and relatively stable fluid flow in the cavity of the heat exchanger, Model thinner wall of Model #2 and relatively stable fluid flow inin thethe cavity of of thethe heat exchanger, Model #2 thinner wall of Model #2 and relatively stable fluid flow cavity heat exchanger, Model #2 was chosen as the carrier to investigate the influence of fluid flow direction on the output power. was chosen as the carrier to investigate thethe influence of fluid flow direction on on thethe output power. In #2 was chosen as the carrier to investigate influence of fluid flow direction output power. In the experiments, we controlled the ∆T from around 40 to◦ 80 °C, held the flow rates through the the experiments, we we controlled the the ∆T ∆T from around 40 to C, held the flow ratesrates through the hot In the experiments, controlled from around 4080to 80 °C, held the flow through the hot and cold sides of the heat exchanger constant, selected Mode 5 as the connection method between and coldcold sides of the heat exchanger constant, selected Mode 5 5asasthe hot and sides of the heat exchanger constant, selected Mode theconnection connectionmethod methodbetween between the thermoelectric modules, and then tested the output power of Model #2 at the different ∆T in the the the thermoelectric thermoelectric modules, modules, and and then then tested tested the the output output power power of of Model Model #2 #2 at at the the different different ∆T ∆T in in the the co-current and counter current. The experimental results are shown in Figure 12. co-current co-current and and counter counter current. current. The The experimental experimentalresults resultsare areshown shownin inFigure Figure12. 12.

Figure 11. Fluid flow directiondiagram: diagram:(a) (a)Co-current; Co-current; (b) (b) Counter Counter current (The dots inin diagram areare Figure Fluid flow direction diagram: (a) Co-current; diagram are Figure 11.11. Fluid flow direction Countercurrent current(The (Thedots dotsin diagram testing points). testing points). testing points). 3030

Countercurrent current Counter Co-current Co-current

25

25

y = 0.3988x - 5.3376 y = 0.3988x - 5.3376 R² = 0.9967 R² = 0.9967

Output Power, W Output Power, W

20 20 15 15

y = 0.3189x - 3.7882 y = 0.3189x - 3.7882 R² = 0.9775

10 10

R² = 0.9775

5

5 0

0

20 20

30 30

40 50 60 70 Difference, ℃ 70 40 Temperature 50 60 Temperature Difference, ℃

80

80

90

90

Figure 12. 12. The The effect effect of of flow flow direction direction on on output output power. power. Figure

Figure 12. The effect of flow direction on output power.

Figure 12 demonstrates that the output power of the counter current is much higher than that of Figure 12 demonstrates that the output power of the counter current is much higher than that theFigure co-current for the same TEG, when ∆T is power the same. Thecounter thermal current loss of the side and the thermal 12 demonstrates that the output of the is hot much higher than that of of the co-current for the same TEG, when ∆T is the same. The thermal loss of the hot side and the absorption of the cold side lead to this phenomenon when the TEG is working. For the hot side, the thethermal co-current for the same when ∆Tto is this the same. The thermal theishot side and thethe thermal absorption of theTEG, cold side lead phenomenon whenloss the of TEG working. For hot temperature at the inlet is higher than that at the outlet, while is opposite for theFor coldthe side. Counter absorption of the cold side phenomenon theit TEG side, the side, the temperature at thelead inlettoisthis higher than that atwhen the outlet, whileisitworking. is opposite for thehot cold side. current may result in a closer temperature difference at the inlet and outlet, but co-current may temperature at themay inletresult is higher at the outlet, while itatisthe opposite foroutlet, the cold Counter Counter current in a than closerthat temperature difference inlet and butside. co-current enlarge thisresult temperature difference (inlet temperature difference is greater than but that co-current at the outlet). current may in a closer temperature difference at the inlet and outlet, may may enlarge this temperature difference (inlet temperature difference is greater than that at the outlet). The output voltage and current is different when the temperature difference of each module is not enlarge this temperature difference (inlet temperature difference is greater than module that at the outlet). The output voltage and current is different when the temperature difference of each is not the the same (as shown in Figure 13), and it leads to more consumption of thermal energy and less output The output voltageinand current is different when temperature difference of each is not same (as shown Figure 13), and it leads to morethe consumption of thermal energy andmodule less output power. Under the same conditions, the output power based on counter current flow mode is greater the same (as shown in Figure 13), and it leads to more consumption of thermal energy and less output than that in the co-current flow mode.

power. Under the same conditions, the output power based on counter current flow mode is greater than that in the co-current flow mode.

Energies 2017, 10, 1329

11 of 15

power. Under the same conditions, the output power based on counter current flow mode is greater Energies 2017, 10, 1329 11 of 15 than that in the co-current flow mode. 100

Temperature, ℃

80

60 Hot side temperature Cold side temperature

40

Temperature difference 20

0 1

2

3 Testing Point

4

5

(a)

100

Temperature, ℃

80

60 Hot side temperature Cold side temperature

40

Temperature difference 20

0 1

2

3 Testing Point

4

5

(b) Figure of each each testing testing point pointfor fordifferent differentflow flowdirection: direction: Figure13. 13.Temperature Temperatureand and temperature temperature difference difference of (a) (a)Co-current; Co-current;(b) (b)Counter Countercurrent. current.

4.6.Effect EffectofofFlow FlowRate Rate 4.6. Afterthe thewall wallof ofthe the fluid fluid channel channel in Model #3 has been After been treated treated to tobe berougher, rougher,the thevelocity velocityofofthe the fluidpassing passingthrough throughthe thefluid fluidchannel channel is is affected. affected. As the influence fluid influence of of the thevolumetric volumetricflow flowrate rateatatthe the inletofofthe theheat heatexchanger exchangeron on the the output output power is likely to be more inlet more distinct, distinct, Model Model#3 #3was waschosen chosenasas thecarrier carriertotostudy studythe therelationship relationship between between the the flow flow rate rate of the inlet and the output the output power power of of TEG. TEG. ◦ In the experiments, ∆T was maintained at a constant 50 °C, the fluid flow direction in the In the experiments, ∆T was maintained at a constant 50 C, the fluid flow direction in theheat heat exchanger plate plate was was counter counter current. hot side inlet flow-rate waswas controlled at 18at exchanger current. On Onone onehand, hand,the the hot side inlet flow-rate controlled L/min unchanged, onlyonly adjusting the cold side side inlet inlet flow-rate from 1from L/min to 14 L/min, and testing 18 L/min unchanged, adjusting the cold flow-rate 1 L/min to 14 L/min, and ◦ C, the output power of Model #2 at each cold side flow-rate. Then ∆T was kept at around 70 °C,70the testing the output power of Model #2 at each cold side flow-rate. Then ∆T was kept at around above experiment was repeated, and thethe experimental results the above experiment was repeated, and experimental resultsare areshown shownininFigure Figure14. 14.On Onthe theother other ◦ hand, we still controlled the ∆T in Model #2 at 50 °C but held the cold side inlet flow-rate constant, hand, we still controlled the ∆T in Model #2 at 50 C but held the cold side inlet flow-rate constant, changing the hot side inlet flow-rate from 1 L/min to 18 L/min, and tested the output power of Model #2 in each flow rate at the hot inlet. The above experiments were repeated while keeping ∆T stabilized at around 70 °C. The experimental results are shown in Figure 15.

Energies 2017, 10, 1329

12 of 15

changing the hot side inlet flow-rate from 1 L/min to 18 L/min, and tested the output power of Model #2 in each flow rate at the hot inlet. The above experiments were repeated while keeping ∆T stabilized Energies 2017, 10, 1329 12 of 15 ◦ C. 2017, 1329 12 of 15 atEnergies around 7010, The experimental results are shown in Figure 15. 35 35


30 30 2525 2020 1515 1010

∆T=70℃ ∆T=70℃

55 00

∆T=50℃ ∆T=50℃ 00

22

44

66 88 10 12 10 12 Cold Fluid Flow Rates, L/min Cold Fluid Flow Rates, L/min

14 14

16 16

Figure Figure14. 14.The Theeffect effectof rate on on output output Figure 14. The effect ofcold coldfluid fluidflow flowrate output power. power. 3535 3030


2525 2020 1515 1010

∆T=70℃ ∆T=70℃

55 00

∆T=50℃ ∆T=50℃

00

22

44

66

10 12 14 88 10 12 14 HotFluid FluidFlow FlowRates, Rates, L/min L/min Hot

16 16

18 18

20 20

Figure15. 15.The Theeffect effectof ofhot hotfluid fluidflow flow rate rate on output power. Figure Figure 15. The effect of hot fluid flow rate on on output output power. power.

The inletflow-rate flow-rateofofthe thehot hot sidewas wasconstant, constant,the theoutput output power power increases increases first and then slowly The first slowly The inlet inlet flow-rate of the hotside side was constant, the output power increases firstand andthen then slowly becomes constant with the increase in the cold side inlet flow-rate. The hot inlet flow-rate is 18 L/min becomes constant with the increase in the cold side inlet flow-rate. The hot inlet flow-rate is 18 L/min becomes constant with the increase in the cold side inlet flow-rate. The hot inlet flow-rate is 18 L/min andthe thecold coldinlet inletflow-rate flow-rateisis88L/min, L/min,the themaximum maximumoutput output power power of of Model Model #3 #3 is 26.8 W under aa ∆T and is 26.8 W under and the cold inlet flow-rate is 8 L/min, the maximum output power of Model #3 is 26.8 W under a ∆T ∆T of 70 °C (see Figure 14). of Figure 14). ◦ C(see of 70 70 °C (see Figure 14). According to Figure 15, the output power also increases first and then tends to be stabilized with According to Figure 15, the power also increases first and then to with According 15,inlet the output output power also then tends tends to be bestabilized stabilized the increase of to theFigure hot side flow-rate when theincreases cold sidefirst inletand flow-rate is unchanged. Whenwith the the increase of the hot side inlet flow-rate when the cold side inlet flow-rate is unchanged. When the the increase of the hot side inlet flow-rate when the cold side inlet flow-rate is unchanged. When the ∆T is 70 °C, the cold side inlet flow is 11.7 L/min and the hot side inlet flow is 7 L/min, the output ∆T is 70 °C, the cold side inlet flow is 11.7 L/min and the hot side inlet flow is 7 L/min, the output ◦ ∆T is 70 C, the cold side inlet flow is 11.7 L/min and the hot side inlet flow is 7 L/min, the output power of Model #3 is maximized around 27.1 W. power of Model is around 27.1 W. powerIn ofsumming Model #3 #3up, is maximized maximized around when the inlet flow27.1 rateW. of the cold (hot) side of the heat exchanger is constant, In summing up, when the inlet flow rate of (hot) side of exchanger isisconstant, summing up, when thehot inlet flow rateincreasing, ofthe thecold coldthe (hot) sidepower ofthe theheat heat constant, withInonly the flow rate at the (cold) side output of theexchanger Model #3 tends to be with only the flow rate at the hot (cold) side increasing, the output power of the Model #3 tends to with only the flow rate at the hot (cold) side increasing, the output power of the Model #3 tends to be be stable. The rise in output power is due to the fact that the heat transfer is improved while the inlet stable. The rise in output power is due to the fact that the heat transfer is improved while the inlet stable. Theincreasing, rise in output is due the fact that the heatthe transfer is improved while the inlet flow was but ifpower the inlet flowtocontinues to increase, heat transfer tends to be balanced flow was increasing, but flow continues to increase, the heat tends to balanced flow increasing, but ifif the theIninlet inlet continues increase, theoptimal heat transfer transfer tends to be be balanced and was reached the maximum. TEGflow devices, there to may exist an flow rate. Fluid flow rates and reached the maximum. In TEG devices, there may exist an optimal flow rate. Fluid flow rates exceeding the optimal value may not bring more power output, but increase the self-consumption. exceeding the optimal value may not bring more power output, but increase the self-consumption.

Energies 2017, 10, 1329

13 of 15

and reached the maximum. In TEG devices, there may exist an optimal flow rate. Fluid flow rates Energies 2017,the 10, 1329 13 of 15 exceeding optimal value may not bring more power output, but increase the self-consumption. Energies 2017, 10, 1329 13 of 15 5. Optimization Results 5. Optimization Results Based on above optimization, an optimized model improved from Model #3 was designed, built, Based on above optimization, an optimized◦ model improved from Model #3 was designed, built, and tested. value of ∆T at 80 C. fact that the more thermoelectric and tested.The The value of was ∆T controlled was controlled at On 80 the °C.basis On of thethebasis of the fact that the more and tested. The value of ∆T was controlled at 80 °C. On the basis of the fact that the more modules in series, the greater thethe total internal consumption, Mode 4 (TableMode 3) was selected thermoelectric modules in series, greater theresistance total internal resistance consumption, 4 (Table 3) thermoelectric modules in series, the greater the total internal resistance consumption, Mode 4 (Table 3) as the connection between thermoelectric modules. The fluid flow direction of the parallel was selected as themethods connection methods between thermoelectric modules. The fluid flow direction of was selected as the connection methods between thermoelectric modules. The fluid flow direction of heat exchange was counter current. The appropriate flow rate flow on the hotonand sides of the parallel heatplates exchange plates was counter current. The appropriate rate the cold hot and cold the parallel heat exchange plates was counter current. The appropriate flow rate on the hot and cold the heat exchanger was applied. The maximum output power average maximum output output power sides of the heat exchanger was applied. The maximum output and power and average maximum sides of the heat exchanger was applied. The maximum output power and average maximum output of Model was tested, the comparison with the previously data [13] is presented power of #3 Model #3 wasand tested, and the comparison with the reported previously reported data [13] in is power of Model #3 was tested, and the comparison with the previously reported data [13] is Figures 16 in and 17. 16 and 17. presented Figures presented in Figures 16 and 17. 35 35 Optimized model Optimized model Previous model Previous model

30 30

OutputPower, Power,WW Output

25 25

y = 0.5516x + 0.58 y = R² 0.5516x + 0.58 = 0.9936 R² = 0.9936

20 20 y = 0.7889x + 0.9019 y = 0.7889x + 0.9019 R² = 0.9978 R² = 0.9978

15 15 10 10 5 5 0 0

0 0

5 5

10 10

15 20 25 30 35 15The Quantity 20 25of Modules 30 35

The Quantity of Modules

40 40

45 45

50 50

Figure 16. Maximum the optimized model and previous model model at different different ∆T. Figure Maximum output output power power of of Figure 16. 16. Maximum output power of the the optimized optimized model model and and previous previous model at at different ∆T. ∆T.

AverageOutput OutputPower, Power,WW Average

1.5 1.5 Optimized model Optimized model

1.2 1.2

Previous model Previous model

0.9 0.9 0.6 0.6 0.3 0.3 0.0 0.0

0 0

5 5

10 10

15 20 25 30 35 15 20 25 30 35 The Quantity of Modules The Quantity of Modules

40 40

45 45

50 50

Figure 17. Average maximum output power of the optimized and previous models at different ∆T. Figure maximum output output power power of of the the optimized optimized and and previous previous models models at at different different ∆T. ∆T. Figure 17. 17. Average Average maximum

As demonstrated in Figures 16 and 17, the output power of the optimized model has been As demonstrated in Figures 16 and 17, the output power of the optimized model has been As demonstrated in Figures 16 and output power of the optimized model enhanced by about 34.6% compared with 17, thethe output power of the previous model [13], has the enhanced by about 34.6% compared with the output power of the previous model [13], the been enhanced by about 34.6% compared with the output power of the previous model [13], instantaneous efficiency of the optimized TEG system could reach about 6.5% at the same instantaneous efficiency of the optimized TEG system could reach about 6.5% at the same the instantaneous efficiency of the optimized TEG system could reach about 6.5% at the same temperature difference. temperature difference. temperature difference.

Energies 2017, 10, 1329

14 of 15

6. Conclusions (1) (2) (3)

The experimental results show that the maximum output power is greatly affected by the operating conditions, especially the TEG apparatus structure, fluid flow direction and flow rates. The maximum output power is not affected by the connection patterns (series or parallel) of the modules, although the voltages may change. A TEG system has been designed, built, tested, and optimized. Experimental studies on the internal resistance of the TEG module and the main factors that affect the output power were conducted. Compared with the previously published system, the output power of the optimized TEG has increased more than 34.6% (from 22.5 W to 32.5 W for 40 TEG modules); the instantaneous efficiency of the TEG system could reach about 6.5%.

Acknowledgments: This research was conducted partially with financial support from China Pingmei Shenma Group, the contributions of which are gratefully acknowledged. Author Contributions: K.L. conceived and designed the experiments; J.C. and C.L. designed and performed the experiments; M.L. and Y.L. contributed materials and tools; J.C., K.L., C.L., L.J. and S.J. conducted the data analysis and wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5.

6. 7.

8. 9.

10. 11. 12. 13. 14.

Wright, P.M. The sustainability of production from geothermal resources. Bull. Geo-Heat Cent. 1998, 19, 9–12. Tang, X.; Zhang, J.; Pang, Z.; Hu, S.; Wu, Y.; Bao, S. Distribution and genesis of the eastern Tibetan Plateau geothermal belt, western China. Environ. Earth Sci. 2017, 76, 31. [CrossRef] Lund, J.W.; Freeston, D.H.; Boyd, T.L. Direct application of geothermal energy: 2005 worldwide review. Geothermics 2005, 34, 691–727. [CrossRef] Lund, J.W.; Bjelm, L.; Bloomquist, G.; Mortensen, A.K. Characteristics, development and utilization of geothermal resources—A Nordic perspective. Episodes 2008, 31, 140–147. Kalina, A.I. Combined cycle and waste heat recovery power systems based on a novel thermodynamic energy cycle utilizing low-temperature heat for power generation. In Proceedings of the 1983 Joint Power Generation Conference, Indianapolis, IN, USA, 25–29 September 1983. Khaliq, A.; Kumar, R.; Dincer, I. Performance analysis of an industrial waste heat-based trigeneration system. Int. J. Energy Res. 2009, 33, 737–744. [CrossRef] Poran, A.; Tartakovsky, L. Performance and emissions of a direct injection internal combustion engine devised for joint operating with high-pressure thermochemical recuperation system. Energy 2017, 124, 214–226. [CrossRef] Pashchenko, D.I. Thermochemical recovery of heat contained in flue gases by means of bioethanol conversion. Therm. Eng. 2013, 60, 438–443. [CrossRef] Poran, A.; Artoul, M.; Sheintuch, M.; Tartakovsky, L. Modeling internal combustion engine with thermo-chemical recuperation of the waste heat by methanol steam reforming. SAE Int. J. Engines 2014, 7, 234–242. [CrossRef] Javani, N.; Dincer, I.; Naterer, G.F. Thermodynamic analysis of waste heat recovery for cooling systems in hybrid and electric vehicles. Energy 2012, 46, 109–116. [CrossRef] Noie-Baghban, S.H.; Majideian, G.R. Waste heat recovery using heat pipe heat exchanger (HPHE) for surgery rooms in hospitals. Appl. Therm. Eng. 2000, 20, 1271–1282. [CrossRef] Li, K.; Bian, H.; Liu, C. Comparison of geothermal with solar and wind power generation systems. Renew. Sustain. Energy Rev. 2015, 42, 1464–1474. [CrossRef] Liu, C.; Chen, P.; Li, K. A 500W low-temperature thermoelectric generator: Design and experimental study. Int. J. Hydrog. Energy 2014, 39, 15497–15505. [CrossRef] Godart, C.; Gonçalves, A.P.; Lopes, E.B.; Villeroy, B. Role of structures on thermal conductivity in thermoelectric materials. In Properties and Applications of Thermoelectric Materials; Veljko, Z., Alex, C.H., Eds.; Springer: Berlin, Germany, 2009; pp. 19–49.

Energies 2017, 10, 1329

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

32. 33.

15 of 15

Zheng, X.F.; Liu, C.X.; Yan, Y.Y. A review of thermoelectric research—Recent developments and potentials for sustainable and renewable energy applications. Renew. Sustain. Energy Rev. 2014, 486–503. [CrossRef] Demir, M.E.; Dincer, I. Development of an integrated hybrid solar thermal power system with thermoelectric generator for desalination and power production. Desalination 2017, 404, 59–71. [CrossRef] Xing, N.; Yu, J.; Wang, S. Experimental study on low-temperature waste heat thermoelectric generator. J. Power Sources 2009, 188, 621–626. Chen, L.; Gong, J.; Sun, F. Effect of heat transfer on the performance of thermoelectric generators. Int. J. Therm. Sci. 2002, 41, 95–99. [CrossRef] Crane, D.T.; Jackson, G.S. Optimization of cross flow heat exchangers for thermoelectric waste heat recovery. Energy Convers. Manag. 2004, 45, 1565–1582. [CrossRef] Hsu, C.T.; Huang, G.Y.; Chu, H.S. Experiments and simulations on low-temperature waste heat harvesting system by thermoelectric power generators. Appl. Energy 2011, 88, 1291–1297. [CrossRef] Rodríguez, A.; Vián, J.G.; Astrain, D. Study of thermoelectric systems applied to electric power generation. Energy Convers. Manag. 2009, 50, 1236–1243. [CrossRef] Bélanger, S.; Gosselin, L. Thermoelectric generator sandwiched in a cross flow heat exchanger with optimal connectivity between modules. Energy Convers. Manag. 2011, 52, 2911–2918. [CrossRef] Ming, T.; Wu, Y.; Peng, C. Thermal analysis on a segmented thermoelectric generator. Energy 2015, 80, 388–399. [CrossRef] Esarte, J.; Min, G.; Rowe, D.M. Modelling heat exchangers for thermoelectric generators. J. Power Sources 2001, 93, 72–76. [CrossRef] Gou, X.; Xiao, H.; Yang, S. Modeling, experimental study and optimization on low-temperature waste heat thermoelectric generator system. Appl. Energy 2010, 87, 3131–3136. [CrossRef] Casano, G.; Piva, S. Experimental investigation of the performance of a thermoelectric generator based on Peltier cells. Exp. Therm. Fluid Sci. 2011, 35, 660–669. [CrossRef] Chen, M.; Rosendahl, L.A.; Condra, T. A three dimensional numerical model of thermoelectric generators in fluid power systems. Int. J. Heat Mass Transf. 2011, 54, 345–355. [CrossRef] Yu, J.; Zhao, H. A numerical model for thermoelectric generator with the parallel-plate heat exchanger. J. Power Sources 2007, 172, 428–434. [CrossRef] Meng, F.; Chen, L.; Sun, F. A numerical model and comparative investigation of a thermoelectric generator with multi-irreversibilities. Fuel Energy Abstr. 2011, 36, 3513–3522. [CrossRef] Gou, X.; Yang, S.; Xiao, H. A dynamic model for thermoelectric generator applied in waste heat recovery. Energy 2013, 52, 201–209. [CrossRef] Twaha, S.; Zhu, J.; Yan, Y. Performance analysis of thermoelectric generator using dc-dc converter with incremental conductance based maximum power point tracking. Energy Sustain. Dev. 2017, 37, 86–98. [CrossRef] Putri, R.I.; Wibowo, S.; Rifa, I.M. Maximum Power Point Tracking for Photovoltaic Using Incremental Conductance Method. Energy Proced. 2015, 68, 22–30. [CrossRef] Seif, S.; Thundat, T.; Cadien, K. Evaluation of efficiency factors and internal resistance of thermoelectric materials. Int. J. Energy Res. 2016, 41, 198–206. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).