Experimental Study on a Thermoelectric Generator for Industrial Waste Heat Recovery Based on a Hexagonal Heat Exchanger

energies

Article Experimental Study on a Thermoelectric Generator for Industrial Waste Heat Recovery Based on a Hexagonal Heat Exchanger

Rui Quan 1,2,* , Tao Li 1, Yousheng Yue 1, Yufang Chang 2 and Baohua Tan 3 1 Hubei Key Laboratory for High-eﬃciency Utilization of Solar Energy and Operation Control of Energy Storage System, Hubei University of Technology, Wuhan 430068, China; [email protected] (T.L.); [email protected] (Y.Y.) 2 Hubei Collaborative Innovation Center for High-eﬃciency Utilization of Solar Energy, Hubei University of Technology, Wuhan 430068, China; [email protected] 3 School of Science, Hubei University of Technology, Wuhan 430068, China; [email protected] * Correspondence: [email protected]; Tel.: +86-27-5975-0858

Received: 22 March 2020; Accepted: 12 June 2020; Published: 17 June 2020

Abstract: To study on the thermoelectric power generation for industrial waste heat recovery applied in a hot-air blower, an experimental thermoelectric generator (TEG) bench with the hexagonal heat exchanger and commercially available Bi2Te3 thermoelectric modules (TEMs) was established, and its performance was analyzed. The influences of several important influencing factors such as heat exchanger material, inlet gas temperature, backpressure, coolant temperature, clamping pressure and external load current on the output power and voltage of the TEG were comparatively tested. Experimental results show that the heat exchanger material, inlet gas temperature, clamping pressure and hot gas backpressure significantly affect the temperature distribution of the hexagonal heat exchanger, the brass hexagonal heat exchanger with lower backpressure and coolant temperature using ice water mixture enhance the temperature difference of TEMs and the overall output performance of TEG. Furthermore, compared with the flat-plate heat exchanger, the designed hexagonal heat exchanger has obvious advantages in temperature uniformity and low backpressure. When the maximum inlet gas temperature is 360 ◦C, the maximum hot side temperature of TEMs is 269.2 ◦C, the maximum clamping pressure of TEMs is 360 kg/m2, the generated maximum output power of TEG is approximately 11.5 W and the corresponding system efficiency is close to 1.0%. The meaningful results provide a good guide for the system optimization of low backpressure and temperature-uniform TEG, and especially demonstrate the promising potential of using brass hexagonal heat exchanger in the automotive exhaust heat recovery without degrading the original performance of internal combustion engine.

Keywords: industrial waste heat recovery; thermoelectric generator; hexagonal heat exchanger; temperature distribution; output performance

1. Introduction The continuously updated development of green energy techniques is a good alternative to resolve the global energy crisis and increase environmental protection. Owing to several advantages, such as little vibration, high reliability and durability and no moving parts, thermoelectric modules (TEMs) have been widely developed in photovoltaic, automotive, military, aerospace, wearable devices, wireless sensor networks and microelectronic applications over the past years [1–9]. Jaziri et al. [1] described and concluded the exploitation of thermoelectric generators (TEGs) in various ﬁelds starting from low-power applications (medical and wearable devices, IoT: internet of things, and WSN: wireless

Energies 2020, 13, 3137; doi:10.3390/en13123137 www.mdpi.com/journal/energies Energies 2020, 13, 3137 2 of 14 sensor network) to high-power applications (industrial electronics, automotive engines and aerospace). Liu et al. [2] developed a TEG based on concentric filament architecture for low power aerospace microelectronic devices, which was able to produce an electrical voltage of 83.5 mVand an electrical power of 32.1 µW with a planar heat source and temperature of 398.15 K. Willars-Rodríguez et al. [3] created and studied a solar hybrid system including photovoltaic (PV) module, concentrating Fresnel lens, thermo-electric generator (TEG) and running water heat extracting unit. Demir et al. [4] proposed a novel system for recovering waste heat of the automobile by a system based thermoelectric generator, and presented the variations of material properties, efficiency and generated power with respect to temperature and position. Meng et al. [5] proposed a technical solution recycling exhaust gas sensible heat based on thermoelectric power generation, which can produce about 1.47 kW electrical energy per square meter and achieve a conversion efficiency of 4.5% for exhaust gas at 350 degrees. Proto et al. [6] analyzed the results of measurements on thermal energy harvesting through a wearable thermoelectric generator (TEG) placed on the arms and legs whose generated power values were in the range from 5 to 50 µW. Kim et al. [7] demonstrated a self-powered wireless sensor node (WSN) driven by a flexible thermoelectric generator (f-TEG) with a significantly improved degree of practicality, and developed a large-area f-TEG that can be wrapped around heat pipes with various diameters, improving their usability and scalability. Leonov et al. [8] studied a thermoelectric energy harvesting on people that generated power in a range of 5–0.5 mW at ambient temperatures of 15 ◦C–27 ◦C, respectively. Holgate et al. [9] conceptualized and modeled a newly designed enhanced multi-mission radioisotope thermoelectric generator (MMRTG) that utilized the more efficient skutterudite-based thermoelectric materials, and presented a discussion of the motivations, modeling results and key design factors. With the rapid development of the economy and society, there is considerable unused waste heat dissipated in industrial heat-generating processes such as power industrial boilers, steelmaking, central air-conditioning, heating equipment and so on. Thermoelectric generator (TEG) is a promising energy source that includes a hot source (heat exchanger) and a cold source (cooling unit); it can convert the heat into electricity based on the temperature difference of the installed TEMs that are sandwiched between the heat exchanger and cooling unit. In the case of TEG for waste heat recovery, many studies have focused on the application and optimization of the flat-plate heat exchanger to obtain higher power output or evaluate its performance [10–14]. For instance, regarding automotive applications, Zhang et al. [10] developed an automotive exhaust thermoelectric generator (AETEG) with 300 TEGs and water cooling for recovering diesel engine exhaust heat, and obtained 1002.6 W power and 2.1% efficiency when the average exhaust temperature was 823 K and the mass flow rate was 480 g/s. Szybist et al. [11] conducted an experiment to investigate 72 TEGs installed on the downstream of three-way catalytic converter and cooled by engine coolant based on a medium-sized gasoline passenger sedan, and 50 W generation power could be generated in the three typical cycles such as FTP (Federal Test Protocol), HWFET (Highway Fuel Economy Test) and US06 cycles. Kim et al. [12] put forward a novel AETEG with heat pipes and measured a maximum power of 350W with 443 K at the evaporator surface of heat pipes. Merkisz et al. [13] placed 24 TEGs at a distance of about 1.5 m from the end of three-way catalytic converter for a gasoline engine and generated the maximum power (189.3W) and maximum efficiency (1.3%) of AETEG, corresponding to the engine speed of 2600 rpm and 2200 rpm, respectively. Fernandez-Yanez et al. [14,15] developed a diesel engine and a gasoline engine AETEGs with engine coolant cooling, and concluded that the maximum produced power was approximately 270 W for both engines if a bypass was not included. However, two non-negligible problems existing in the flat-plate heat exchanger are the heat uniformity and the unwanted backpressure. The former plays an important role in the average temperature difference and thermoelectric efficiency of TEGs, as it dominates the energy conversion efficiency of heat to electricity. The latter reduces the efficiency of the engine for it seriously affects the original fuel economy and emission performance, or even worse, the lost performance of the engine cannot be compensated by the small generated power of TEG. Therefore, an ideal heat exchanger used in a TEG especially in an AETEG, should provide sufficient Energies 2020, 13, 3137 3 of 14 Energies 2020, 13, x FOR PEER REVIEW 3 of 15 uniformity,surface area make to install the TEMs backpressure as much as as low possible, as possible increase and its ensure surface the temperature waste heat uniformity, gas flow makeat a high the value.backpressure as low as possible and ensure the waste heat gas flow at a high value. In this paper, we expanded the previous study to develop a low low-cost‐cost and symmetrical symmetrical TEG based on a brass hexagonal heat exchanger and the commercially available BiBi22Te33 TEMs to recover the waste heat from a hot-airhot‐air blower, which can simulate the internal combustion engine engine used used in in the the AETEG. AETEG. This study aimed to validate the promising potential and advantages of using the hexagonal heat exchanger in waste heat recovery, and analyze the influences influences of several important influencing influencing factors on the output performance of the TEG. This work can provide an application guideline for industrial heatheat-generating‐generating processes and automotive exhaust heatheat recovery.recovery.

2. Experimental Setup of a TEG System A detailed schematic diagram diagram of of the the designed designed TEG TEG system system is is shown shown in in Figure Figure 1;1 ;it it consists consists of of a hota hot-air‐air blower, blower, a ahexagonal hexagonal heat heat exchanger, exchanger, a apump, pump, a a radiator, radiator, a a water water tank, tank, a a pressure regulator, temperature and and pressure pressure sensors sensors (P (P1 1denotesdenotes pressure pressure sensor, sensor, T1T denotes1 denotes inlet inlet temperature temperature sensor, sensor, T2 Tdenotes2 denotes outlet outlet temperature temperature sensor sensor)) and and six six groups groups of of TEMs TEMs and and cooling cooling units. units. The The hot hot-air‐air blower provides wastewaste heatheat toto the the hexagonal hexagonal heat heat exchanger, exchanger, which which can can also also simulate simulate di differentfferent driving driving cycles cycles of vehiclesof vehicles by adjustingby adjusting the gasthe temperature,gas temperature, flow andflow pressure. and pressure. When theWhen hot-air the blowerhot‐air works,blower exhaustworks, exhaustgas flows gas into flows the hexagonal into the hexagonal heat exchanger heat toexchanger provide theto provide hot side temperature,the hot side temperature, and the cooling and water the (i.e.,cooling coolant) water stored (i.e., coolant) in the water stored tank in isthe circulated water tank with is thecirculated pump amongwith the the pump cooling among units the to form cooling the unitscold side.to form Therefore, the cold side. electricity Therefore, is generated electricity due is generated to the temperature due to the ditemperaturefference between difference the between hot side theand hot cold side side and of cold each side TEM. of each To obtain TEM. higher To obtain temperature higher temperature difference difference and better and performance, better performance, the rotate thespeed rotate of radiator speed of can radiator be controlled can be controlled to precool to the precool inlet coolantthe inlet in coolant cooling in units. cooling units.

Figure 1. Schematic of an experimental setup of a TEG System. TEM: thermoelectric module. Figure 1. Schematic of an experimental setup of a TEG System. TEM: thermoelectric module. The specific architecture of TEG and the dimensions of the hexagonal heat exchanger are shown in The specific architecture of TEG and the dimensions of the hexagonal heat exchanger are shown Figure2. Each group of TEMs is sandwiched between the surface of the hexagonal heat exchanger and in Figure 2. Each group of TEMs is sandwiched between the surface of the hexagonal heat exchanger a cooling unit, and they are clamped with an adjustable force using five bolt and screw combinations. and a cooling unit, and they are clamped with an adjustable force using five bolt and screw In total, there are 30 TEMs of Bi2Te3-based materials arranged on the six surfaces of the heat exchanger combinations. In total, there are 30 TEMs of Bi2Te3‐based materials arranged on the six surfaces of the in five columns (i.e., five TEMs in each column are fixed with a common cooling box), and all the heat exchanger in five columns (i.e., five TEMs in each column are fixed with a common cooling box), TEMs stalled above the surface of the hexagonal heat exchanger are electrically connected in series. and all the TEMs stalled above the surface of the hexagonal heat exchanger are electrically connected Furthermore, to guarantee uniform cold side temperature of TEMs, all the cooling boxes are thermally in series. Furthermore, to guarantee uniform cold side temperature of TEMs, all the cooling boxes are connected in parallel (i.e., all the inlets of six cooling boxes are connected with the outlet of radiator, thermally connected in parallel (i.e., all the inlets of six cooling boxes are connected with the outlet of all the outlets of six cooling boxes are connected with the inlet of water tank). To evaluate the hot radiator, all the outlets of six cooling boxes are connected with the inlet of water tank). To evaluate side temperatures distribution of the 30 TEMs, the corresponding K-type thermocouples below each the hot side temperatures distribution of the 30 TEMs, the corresponding K‐type thermocouples TEM are pasted above each surface of the hexagonal heat exchanger. For the TEG, the Bi2Te3 TEMs below each TEM are pasted above each surface of the hexagonal heat exchanger. For the TEG, the (Model Name: TEHP1-1264-0.8) are manufactured by Thermonamic Electronics (Jiangxi, China) Corp. Bi2Te3 TEMs (Model Name: TEHP1‐1264‐0.8) are manufactured by Thermonamic Electronics (Jiangxi, Ltd., whose high conductivity graphite paper is used as the thermally conductive interface material. China) Corp. Ltd., whose high conductivity graphite paper is used as the thermally conductive The specific parameters of the principal components of the TEG system are provided in Table1. interface material. The specific parameters of the principal components of the TEG system are provided in Table 1.

Energies 2020, 13, x FOR PEER REVIEW 4 of 15

Energies 2020, 13, 3137 4 of 14 mm Energies 2020, 13, x FOR PEER REVIEW 4 of 15 53

Figure 2. TEG architecture. mm

Table 1. Parameters of the principal53 components of a TEG system.

Parameters Value Dimension of TEM 40 mm × 40 mm × 40 mm Figure 2. TEG architecture. Maximum operation temperatureFigure 2. TEG of architecture.TEM 400 ℃ Table 1. Parameters of the principal components of a TEG system.℃ Rated operationTable 1. Parameters temperature of the principal of TEM components of a TEG330 system. Maximum conversionParameters efficiency of TEM Value 6.5% Parameters Value DimensionDimension of cooling of TEM box 40250 mm mm40 mm × 5040 mm mm × 20 mm Dimension of TEM 40 mm× × 40 mm× × 40 mm Maximum operation temperature of TEM 400 ◦C MaximumMaterial operation of cooling temperature box of TEM 400Aluminum ℃ Rated operation temperature of TEM 330 ◦C MaximumRatedThickness operation conversion of cooling temperature efficiency box ofof TEM TEM 6.5%330 1℃ mm MaximumDimension conversion of cooling efficiency box of TEM 250 mm 50 mm6.5% 20 mm Thickness of heat exchanger × 2× mm DimensionMaterial of of cooling cooling box box 250 mm Aluminum × 50 mm × 20 mm MaximumThickness power of cooling of pump box 1 mm 40 W Material of cooling box Aluminum MaximumThickness flow of heat of exchanger pump 2 mm5000 L/H MaximumThickness power of cooling of pump box 401 W mm RatedThicknessMaximum power of flowof heat radiator of exchanger pump 50001002 L mmW/H (24V DC) Rated powerMaximumRated powerof hotpower of‐air radiator of blower pump 100 W (24V402000 W DC) W RatedMaximum power of flow hot-air of pump blower 20005000 W L/H The real experimental setupRated of power a TEG of systemradiator is shown in Figure100 W 3;(24V the DC) output of TEG is connected with an adjustableThe real electronic experimentalRated load, setup power and of of a the TEGhot ‐experimentsair system blower is shown described in Figure3 ;in2000 the the output W next of section TEG is connected were carried out with an adjustable electronic load, and the experiments described in the next section were carried out to to evaluate several important influencing factors such as heat exchanger material, inlet gas evaluateThe several real experimental important influencingsetup of a TEG factors system such is asshown heat exchangerin Figure 3; material, the output inlet of TEG gas temperature, is connected temperature,backpressure,with anbackpressure, adjustable coolant electronic temperature coolant load, andtemperature and the external experiments load and currentdescribed external on in its theload temperature next current section distribution were on carriedits temperature and out distributionoutputto evaluate and performance. output several performance. important influencing factors such as heat exchanger material, inlet gas temperature, backpressure, coolant temperature and external load current on its temperature distribution and output performance.

Figure 3. Real experimental setup of a TEG system. FigureFigure 3. Real 3. Real experimental experimental setupsetup of of a TEGa TEG system. system. 3. Experimental Results and Discussions 3. Experimental Results and Discussions 3.1. Temperature Distribution

3.1. TemperatureTo analyze Distribution the temperature uniformity of the hexagonal heat exchanger, 30 independent K‐type thermocouples pasted above its six surfaces are divided into five columns, and each K‐type To analyze the temperature uniformity of the hexagonal heat exchanger, 30 independent K‐type thermocouples pasted above its six surfaces are divided into five columns, and each K‐type

Energies 2020, 13, 3137 5 of 14

3. Experimental Results and Discussions Energies 2020, 13, x FOR PEER REVIEW 5 of 15 3.1. Temperature Distribution thermocouple is installed right below the central hot side of the corresponding single TEM. The To analyze the temperature uniformity of the hexagonal heat exchanger, 30 independent specificK-type surface thermocouples temperature pasted distribution above its of six the surfaces 30 temperature are divided into detected five columns, locations and corresponding each K-type to the 30 TEMsthermocouple is shown is installed in Figure right 4. below On this the occasion, central hot the side hexagonal of the corresponding heat exchanger single TEM. is made The specific of stainless steel (Modelsurface Name: temperature 304), distribution its inlet temperature of the 30 temperature is 260 °C, detected the pressure locations corresponding regulator is tofully the 30open TEMs and the effectiveis showntransfer in size Figure of 4its. Oninterior this occasion, fins is 260 the mm hexagonal in length heat and exchanger 42 mm in is width made of(1.5 stainless mm in steel thickness). For the(Model interior Name: cavity 304), of itsheat inlet exchange temperature is much is 260 ◦ largerC, the pressure than the regulator inlet tube, is fully the open gas andflows the quickly effective across transfer size of its interior fins is 260 mm in length and 42 mm in width (1.5 mm in thickness). For the the inlet section. Furthermore, the gas flow decreases accordingly from column 1 to column 5 because interior cavity of heat exchange is much larger than the inlet tube, the gas flows quickly across the of the pressureinlet section. caused Furthermore, by the interior the gas flow fins, decreases which enhances accordingly the from heat column transfer 1 to between column 5 becausethe hot ofgas and heat exchanger.the pressure Thus, caused it bycan the be interior concluded fins, which that the enhances section the closer heat transfer to the betweeninlet has the the hot lowest gas and surface temperature,heat exchanger. the temperatures Thus, it can of be the concluded detected that locations the section marked closer with to the blue inlet near has the the lowest exhaust surface gas outlet is highertemperature, than those the in temperatures the central of area the detected and closer locations to the marked exhaust with gas blue inlet near thebecause exhaust of gas the outlet gas eddy causedis by higher the sudden than those narrow in the centraloutlet areatube and and closer the temperatures to the exhaust gas of inletthe six because detected of the locations gas eddy in the same columncaused by are the almost sudden the narrow same outlet (the maximum tube and the temperature temperatures is of below the six detected2 °C). Thus, locations compared in the with same column are almost the same (the maximum temperature is below 2 C). Thus, compared with our previously designed chaos‐shaped flat‐plate heat exchanger [16,17],◦ the temperature uniformity our previously designed chaos-shaped flat-plate heat exchanger [16,17], the temperature uniformity of of the hexagonalthe hexagonal heat heat exchanger exchanger is better,better, which which is inis goodin good agreement agreement with our with expected our expected results. results.

FigureFigure 4. The 4. surfaceThe surface temperature temperature distribution distribution of of the the hexagonal hexagonal heat heat exchanger. exchanger. 3.2. Influence of Heat Exchanger Material 3.2. Influence of Heat Exchanger Material For the above designed hexagonal heat exchanger shown in Figure2, two kinds of di fferent Formetal the materials above designed are adopted hexagonal without changing heat exchanger its dimension shown to analyze in Figure their influences 2, two onkinds the surface of different metal materialstemperature are distribution adopted andwithout the output changing performance its dimension of TEG system. to analyze The first their one influences is made of the on above the surface temperaturestainless distribution steel (Model and Name: the 304), output and theperformance other one is madeof TEG of brass.system. The The compared first one characteristic is made of the of TEG with the two different kinds of hexagonal heat exchangers is shown in Figure5. In this case, above stainless steel (Model Name: 304), and the other one is made of brass. The compared the clamping pressure of TEMs above each surface increases is 120 kg/m2, the inlet gas pressure is characteristic70 Pa, the of coolant TEG with is pumped the two without different radiator, kinds the coolantof hexagonal flow is 5000 heat L exchangers/h and its original is shown temperature in Figure 5. 2 In this iscase, equal the to theclamping ambient pressure temperature of (27TEMs◦C). above each surface increases is 120 kg/m , the inlet gas pressure is 70 Pa, the coolant is pumped without radiator, the coolant flow is 5000 L/h and its original temperature is equal to the ambient temperature (27 °C). Figure 5a,b demonstrate the surface temperature distribution of the two heat exchangers with different inlet temperatures of 155 °C, 200 °C, 260 °C, 315 °C and 360 °C, respectively. Considering the above uniform temperature distribution in each column shown in Figure 4, only the temperature detected locations in the NO.1 surface are selected as the average temperatures of the hexagonal heat exchangers (i.e., the hot side temperatures of TEMs) in each column. It is obvious that the average temperatures of both the stainless steel and brass hexagonal heat exchangers are in direct proportion to the inlet gas temperature. According to the Fourier equation, the absorbed heat (denoted Q) of heat exchanger can be calculated as follows:

QKATd=/ (1)

Energies 2020, 13, x FOR PEER REVIEW 6 of 15 where K is the heat conductivity coefficient, A is the heat transfer area, ΔT is the temperature variation and d is the heat transfer distance. For the average heat conductivity of stainless steel 304 is about 16.2 W/m.K, while the one of brass is about 120 W/m.K, which is almost seven times of the former. The detected locations temperatures in the same column validated that the brass hexagonal heat exchanger is much better than the stainless steel hexagonal heat exchanger in heat transfer with the same inlet gas temperature. Properly speaking, when the maximum inlet gas temperature is 360 °C, the maximum temperature of the stainless steel hexagonal heat exchanger (in column 5) is 222.1 °C, while the one of the brass hexagonal heat exchanger (in column 5) is 269.2 °C, which is an increase of 21.2%. According to Equation (1), it can be calculated that the absorbed heat of brass heat exchanger is increased by about 930% because of its higher heat conductivity coefficient. Figure 5c shows the corresponding open‐circuit voltage of TEG with the above two kinds of hexagonal heat exchangers, the open‐circuit voltage of TEG based on the two different hexagonal heat exchangers increases with the augment of its inlet gas temperature. Figure 5d provides the measured characteristics curves of voltage and power versus current of TEG with two different hexagonal heat exchangers when the maximum inlet gas temperature is 360 °C. It can be seen that the open‐circuit voltage of TEG with the stainless steel hexagonal heat exchanger is only 19.5 V, while the one with the brass hexagonal heat exchanger is 41.6 V, which is an increase of 113.3%. In addition, the maximum power of TEG with the stainless steel hexagonal heat exchanger is only 1.65 W when the current is 0.15 A, while the one with the brass hexagonal heat exchanger is 7.79 W, which is an increase of nearly four times. For the above coolant of ambient temperature is adopted, the cold side temperatures of TEMs in the same column of both the two kinds of different hexagonal heat exchangers can be seen similar. Furthermore, as shown in Figure 5b, the TEMs installed in the same column above the brass hexagonal heat exchanger have much higher hot side temperatures with the same inlet gas temperatures, it can be concluded that the temperature difference of TEMs with the brass hexagonal heat exchanger is much larger than that with the stainless steel hexagonal heat exchanger for more hot gas heat is absorbed on the same occasion because of the higher heat transfer coefficient of brass. Thus, the brass hexagonal heat exchanger has overwhelming heat‐conducting property advantage over the stainless steel hexagonal heat exchanger despite its high cost, and the TEG system with the Energiesbrass hexagonal2020, 13, 3137 heat exchanger has a better output performance. 6 of 14

Energies 2020, 13, x FOR PEER REVIEW 7 of 15 (a) (b)

FigureFigure 5.5. TheThe comparedcompared characteristiccharacteristic performance of TEG with didifferentﬀerent heat exchangerexchanger materials. ((aa)) SurfaceSurface temperaturetemperature distributiondistribution ofof thethe stainlessstainless steelsteel heatheat exchanger.exchanger. ((bb)) SurfaceSurface temperaturetemperature distributiondistribution ofof thethe brassbrass heatheat exchanger.exchanger. (c) Open-circuitOpen‐circuit voltage of TEG with inlet gas temperatures. ((dd).). VoltageVoltage andand powerpower versusversus current current when when the the maximum maximum inlet inlet temperature temperature is is 360 360◦ °C.C.

3.3. InfluenceFigure5a,b of Backpressure demonstrate the surface temperature distribution of the two heat exchangers with diﬀerent inlet temperatures of 155 ◦C, 200 ◦C, 260 ◦C, 315 ◦C and 360 ◦C, respectively. Considering the aboveConsidering uniform the temperature advantage distribution of the above in each brass column hexagonal shown heat in Figure exchanger4, only used the temperaturein TEG, the detectedinfluence locations of inlet gas in the backpressure NO.1 surface on areTEG selected based on as the averagebrass hexagonal temperatures heat exchanger of the hexagonal is shown heat in 2 exchangersFigure 6. On (i.e., this the occasion, hot side the temperatures clamping pressure of TEMs) of TEMs in each above column. each surface It is obvious increases that is the 120 average kg/m , temperaturesthe coolant is of pumped both the without stainless radiator, steel and the brass coolant hexagonal flow is heat 5000 exchangers L/h and its are original in direct temperature proportion tois theequal inlet to gasthe temperature.ambient temperature According (27 to °C), the Fourierthe inlet equation, gas backpressure the absorbed is adjusted heat (denoted by theQ regulator) of heat exchangeropening. For can the be calculated detected temperature as follows: locations in the 5th column of the brass hexagonal heat exchanger have the highest temperature values, Figure 6a,b show the maximum surface temperatures of the brass hexagonal heat exchanger and theQ = openKA∆‐circuitT/d voltage of TEG corresponding to different(1) inlet gas temperatures with different inlet gas backpressures of 70 Pa, 180 Pa and 220 Pa, respectively. whereFigureK 6c,dis the show heat the conductivity voltage and coe outputﬃcient, powerA is versus the heat current transfer characteristics area, ∆T is the with temperature the above variation different andinletd gasis the backpressures heat transfer when distance. the maximum For the average inlet gas heat temperature conductivity is 360 of °C. stainless steel 304 is about 16.2 W/m.K, while the one of brass is about 120 W/m.K, which is almost seven times of the former. The detected locations temperatures in the same column validated that the brass hexagonal heat exchanger is much better than the stainless steel hexagonal heat exchanger in heat transfer with the same inlet gas temperature. Properly speaking, when the maximum inlet gas temperature is 360 ◦C, the maximum temperature of the stainless steel hexagonal heat exchanger (in column 5) is 222.1 ◦C, while the one of the brass hexagonal heat exchanger (in column 5) is 269.2 ◦C, which is an increase of

(a) (b)

Energies 2020, 13, 3137 7 of 14

21.2%. According to Equation (1), it can be calculated that the absorbed heat of brass heat exchanger is increased by about 930% because of its higher heat conductivity coefficient. Figure5c shows the corresponding open-circuit voltage of TEG with the above two kinds of hexagonal heat exchangers, the open-circuit voltage of TEG based on the two different hexagonal heat exchangers increases with the augment of its inlet gas temperature. Figure5d provides the measured characteristics curves of voltage and power versus current of TEG with two different hexagonal heat exchangers when the maximum inlet gas temperature is 360 ◦C. It can be seen that the open-circuit voltage of TEG with the stainless steel hexagonal heat exchanger is only 19.5 V, while the one with the brass hexagonal heat exchanger is 41.6 V, which is an increase of 113.3%. In addition, the maximum power of TEG with the stainless steel hexagonal heat exchanger is only 1.65 W when the current is 0.15 A, while the one with the brass hexagonal heat exchanger is 7.79 W, which is an increase of nearly four times. For the above coolant of ambient temperature is adopted, the cold side temperatures of TEMs in the same column of both the two kinds of different hexagonal heat exchangers can be seen similar. Furthermore, as shown in Figure5b, the TEMs installed in the same column above the brass hexagonal heat exchanger have much higher hot side temperatures with the same inlet gas temperatures, it can be concluded that the temperature difference of TEMs with the brass hexagonal heat exchanger is much larger than that with the stainless steel hexagonal heat exchanger for more hot gas heat is absorbed on the same occasion because of the higher heat transfer coefficient of brass. Thus, the brass hexagonal heat exchanger has overwhelming heat-conducting property advantage over the stainless steel hexagonal heat exchanger despite its high cost, and the TEG system with the brass hexagonal heat exchanger has a better output performance.

3.3. Influence of Backpressure Considering the advantage of the above brass hexagonal heat exchanger used in TEG, the influence of inlet gas backpressure on TEG based on the brass hexagonal heat exchanger is shown in Figure6. On this occasion, the clamping pressure of TEMs above each surface increases is 120 kg/m2, the coolant is pumped without radiator, the coolant flow is 5000 L/h and its original temperature is equal to the ambient temperature (27 ◦C), the inlet gas backpressure is adjusted by the regulator opening. For the detected temperature locations in the 5th column of the brass hexagonal heat exchanger have the highest temperature values, Figure6a,b show the maximum surface temperatures of the brass hexagonal heat exchanger and the open-circuit voltage of TEG corresponding to different inlet gas temperatures with different inlet gas backpressures of 70 Pa, 180 Pa and 220 Pa, respectively. Figure6c,d show the voltage and output power versus current characteristics with the above different inlet gas backpressures when the maximum inlet gas temperature is 360 ◦C. Larger inlet gas backpressure means lower maximum surface temperature and open-circuit voltage with the same inlet gas temperature; the reason of this is that large inlet gas backpressure will decrease the rotation speed of hot-air blower’s fan and the corresponding flow of hot gas. In this case, the heat transfer between hot gas and heat exchanger is reduced, which leads to a lower temperature difference of TEMs based on the same cooling condition. When the maximum inlet gas temperature is 360 ◦C, the maximum surface temperature of the brass hexagonal heat exchanger is 268.2 ◦C (70 Pa), 250.1 ◦C (180 Pa) and 241.3 ◦C (220 Pa); the open-circuit of TEG is 41.6 V (70 Pa), 35.1 V (180 Pa) and 29.9 V (220 Pa). Furthermore, the open-circuit voltage and output power of TEG are in inverse proportion to the inlet gas backpressures, decreased both output voltage and power with the same output current are accompanied with the augment of inlet gas backpressures. Since the cold side temperature of TEMs in each column with different inlet gas backpressures can be regarded as the same because of the ambient temperature coolant, it can be concluded that the lower inlet gas backpressure is, the larger inlet gas flow will be. Lower inlet gas backpressure contributes to higher temperature differences and better output performance of TEMs, since it ensures higher rotating speed of hot-air Energies 2020, 13, x FOR PEER REVIEW 7 of 15

Figure 5. The compared characteristic performance of TEG with different heat exchanger materials. (a) Surface temperature distribution of the stainless steel heat exchanger. (b) Surface temperature distribution of the brass heat exchanger. (c) Open‐circuit voltage of TEG with inlet gas temperatures. (d). Voltage and power versus current when the maximum inlet temperature is 360 °C.

3.3. Influence of Backpressure Considering the advantage of the above brass hexagonal heat exchanger used in TEG, the influence of inlet gas backpressure on TEG based on the brass hexagonal heat exchanger is shown in Figure 6. On this occasion, the clamping pressure of TEMs above each surface increases is 120 kg/m2, the coolant is pumped without radiator, the coolant flow is 5000 L/h and its original temperature is equal to the ambient temperature (27 °C), the inlet gas backpressure is adjusted by the regulator opening. For the detected temperature locations in the 5th column of the brass hexagonal heat exchanger have the highest temperature values, Figure 6a,b show the maximum surface temperatures of the brass hexagonal heat exchanger and the open‐circuit voltage of TEG corresponding to different Energies 2020, 13, 3137 8 of 14 inlet gas temperatures with different inlet gas backpressures of 70 Pa, 180 Pa and 220 Pa, respectively. Figure 6c,d show the voltage and output power versus current characteristics with the above different blower,inlet gas and backpressures eﬃcient heat when conduction the maximum between inlet the gas gas temperature and brass hexagonalis 360 °C. heat exchanger on the same occasion.

Energies 2020, 13, x FOR PEER REVIEW 8 of 15 (a) (b)

Figure 6.6. TheThe compared compared characteristic characteristic performance performance of TEG of basedTEG onbased the brasson the hexagonal brass hexagonal heat exchanger heat withexchanger diﬀerent with inlet different gas backpressures. inlet gas backpressures. (a) Maximum (a) Maximum surface temperatures surface temperatures of the brass of heat the exchangerbrass heat withexchanger inlet gas with temperatures. inlet gas temperatures. (b) Open-circuit (b) Open voltage‐circuit of TEGvoltage with of inlet TEG gas with temperatures. inlet gas temperatures. (c) Voltage (versusc) Voltage current versus when current the maximum when the maximum inlet temperature inlet temperature is 360 ◦C. (isd )360 Power °C. ( versusd) Power current versus when current the whenmaximum the maximum inlet temperature inlet temperature is 360 ◦C. is 360 °C.

3.4. InfluenceLarger inlet of Coolant gas backpressure Temperatures means lower maximum surface temperature and open‐circuit voltageTo ensurewith the lower same cold inlet side gas temperatures temperature; and the higherreason temperature of this is that di largefferences inlet of gas TEMs backpressure with the same will decreasehot side temperatures, the rotation speed three of di ffhoterent‐air kindsblower’s of coolant fan and were the corresponding used in the test flow as follows: of hot gas. case In 1: this pumped case, thecoolant heat of transfer ambient between temperature hot gas without and heat radiator; exchanger case 2:is reduced, pumped coolantwhich leads of ambient to a lower temperature temperature with differenceradiator of of 100% TEMs speed; based case on 3: the pumped same cooling coolant condition. of ice water When mixture the maximum (0 ◦C) without inlet radiator. gas temperature Figure7 isshows 360 °C, the the characteristic maximum performance surface temperature of TEG based of the on brass the brass hexagonal hexagonal heat heat exchanger exchanger is 268.2 in the °C above (70 Pa),three 250.1 cases °C when (180 the Pa) clamping and 241.3 pressure °C (220 ofPa); TEMs the open above‐circuit each surface of TEG increases is 41.6 V is (70 120 Pa), kg /35.1m2, theV (180 coolant Pa) andflow 29.9 is 5000 V (220 L/h, Pa). the ambient Furthermore, temperature the open is 27‐circuit◦C and voltage the inlet and gas output backpressure power of is maintainedTEG are in atinverse 70 Pa. proportion to the inlet gas backpressures, decreased both output voltage and power with the same output current are accompanied with the augment of inlet gas backpressures. Since the cold side temperature of TEMs in each column with different inlet gas backpressures can be regarded as the same because of the ambient temperature coolant, it can be concluded that the lower inlet gas backpressure is, the larger inlet gas flow will be. Lower inlet gas backpressure contributes to higher temperature differences and better output performance of TEMs, since it ensures higher rotating speed of hot‐air blower, and efficient heat conduction between the gas and brass hexagonal heat exchanger on the same occasion.

3.4. Influence of Coolant Temperatures To ensure lower cold side temperatures and higher temperature differences of TEMs with the same hot side temperatures, three different kinds of coolant were used in the test as follows: case 1: pumped coolant of ambient temperature without radiator; case 2: pumped coolant of ambient temperature with radiator of 100% speed; case 3: pumped coolant of ice water mixture (0 °C) without radiator. Figure 7 shows the characteristic performance of TEG based on the brass hexagonal heat exchanger in the above three cases when the clamping pressure of TEMs above each surface increases is 120 kg/m2, the coolant flow is 5000 L/h, the ambient temperature is 27 °C and the inlet gas backpressure is maintained at 70 Pa.

EnergiesEnergies 20202020,, 1313,, 3137x FOR PEER REVIEW 99 of 1415

(a) (b)

Figure 7.7.The The compared compared characteristic characteristic performance performance of TEG of basedTEG onbased the brasson the hexagonal brass hexagonal heat exchanger heat inexchanger different in coolant different cases. coolant (a) Coolant cases. temperatures(a) Coolant temperatures of TEG with inletof TEG gas with temperatures. inlet gas temperatures. (b) Open-circuit (b) voltageOpen‐circuit of TEG voltage with inletof TEG gas with temperatures. inlet gas (temperatures.c) Voltage versus (c) Voltage current whenversus the current maximum when inlet the temperaturemaximum inlet is 360 temperature◦C. (d) Power is 360 versus °C. (d) current Power whenversus the current maximum when inletthe maximum temperature inlet is temperature 360 ◦C. is 360 °C. For Figure7a, the inlet gas temperatures are changed every five minutes to ensure the steady heat conductionFor Figure among 7a, TEMs, the inlet heat gas exchanger temperatures and cooling are changed boxes. Theevery coolant five minutes temperatures to ensure rise asthe the steady inlet gasheat temperatures conduction among increase, TEMs, the coolant heat exchanger temperature and in cooling case 1 isboxes. the highest, The coolant the coolant temperatures temperature rise inas casethe inlet 2 takes gas second temperatures place, while increase, the coolant the coolant temperature temperature in case in 3 iscase the 1 lowest. is the Whenhighest, the the maximum coolant inlettemperature gas temperature in case 2 istakes 360 second◦C, the maximumplace, while coolant the coolant temperature temperature in case in 1 case is 45.9 3 is◦ C,the the lowest. one in When case 2the is maximum 36.6 ◦C and inlet the gas one temperature in case 3 is 20.2is 360◦ C.°C, Figure the maximum7b shows coolant the open-circuit temperature voltage in case of 1 TEG is 45.9 with °C, ditheff erentone in inlet case gas 2 is temperatures 36.6 °C and the based one on in the case above 3 is 20.2 three °C. di ffFigureerent kinds7b shows of coolant. the open It‐ iscircuit obvious voltage that theof TEG open-circuit with different voltage inlet of TEG gas intemperatures case 3 is the based largest, on the the one above in in three case 2different takes second kinds place, of coolant. while It the is oneobvious in case that 1 the is the open lowest‐circuit with voltage the same of TEG inlet in gas case temperature. 3 is the largest, Figure the one7c,d in show in case the 2 performance takes second characteristicsplace, while the of one TEG in in case the 1 above is the threelowest cases with with the same different inlet load gas currentstemperature. when Figure the maximum 7c,d show inlet the gasperformance temperature characteristics is 360 ◦C. It canof TEG be seen in the that above both the three output cases voltage with different and power load of TEG currents corresponding when the tomaximum the same inlet load gas current temperature increase evidentlyis 360 °C. withIt can lower be seen coolant that both temperatures, the output the voltage maximum and power of TEGTEG incorresponding case 1, 2 and 3to is the 7.79 same W, 8.56 load W current and 9.65 increase W, respectively. evidently Additionally,with lower coolant the coolant temperatures, temperature the willmaximum increase power evidently of TEG without in case a radiator1, 2 and because3 is 7.79 ofW, the 8.56 large W and thermal 9.65 conductivityW, respectively. of TEMs, Additionally, and the coolantthe coolant temperature temperature with will radiator increase can beevidently maintained without in a a relatively radiator stablebecause range. of the To large ensure thermal larger outputconductivity power of and TEMs, higher and output the coolant voltage, temperature a coolant of icewith water radiator mixture can is be recommended maintained in in a the relatively cooling unitstable of range. TEG, asTo it ensure can evidently larger output enlarge power the temperature and higher output difference voltage, of TEM a coolant without of ice consuming water mixture extra radiatoris recommended power. in the cooling unit of TEG, as it can evidently enlarge the temperature difference of TEM without consuming extra radiator power.

Energies 2020,, 13,, 3137x FOR PEER REVIEW 10 of 15 14

3.5. Influence of Clamping Pressure 3.5. Influence of Clamping Pressure Considering the thermal contact resistance caused by the clamping pressure may affect the hot side Consideringand cold side the temperatures thermal contact of TEG, resistance the effect caused of different by the clamping clamping pressure pressures may on aff theect theoverall hot sideperformance and cold of side TEG temperatures is shown in Figure of TEG, 8. On the this effect occasion, of different the coolant clamping of ice pressures water mixture on the (0 overall °C) is performancepumped without of TEG radiator, is shown the in coolant Figure flow8. On is this 5000 occasion, L/h, the the ambient coolant temperature of ice water is mixture 27 °C, (0the◦ C)inlet is pumpedgas backpressure without radiator, is maintained the coolant at 70 flowPa and is 5000 the Lmaximum/h, the ambient inlet temperature isof 27 heat◦C, exchanger the inlet gas is backpressure360 °C. The maximum is maintained output at power 70 Pa and of TEG the maximum is 9.65 W, 10.32 inlet temperatureW and 11.49 ofW heat with exchanger different clamping is 360 ◦C. Thepressures maximum of 120 output kg/m power2, 240 of kg/m TEG2 isand 9.65 360 W, 10.32kg/m2 W, respectively. and 11.49 W withThe dicorrespondingfferent clamping open pressures‐circuit ofvoltage 120 kg of/m TEG2, 240 is kg46.2/m V,2 and 49.1 360 V and kg/m 53.32, respectively. V, respectively. The corresponding open-circuit voltage of TEG is 46.2 V, 49.1 V and 53.3 V, respectively.

(a) (b)

Figure 8. TEGTEG performance performance with clamping pressure ofof 120120 kgkg/m/m22,, 240 kgkg/m/m22 and 360 kgkg/m/m22..( (aa)) Voltage Voltage versus current curves, (b)) power versus currentcurrent curves.curves.

From the separate characteristic curves shown in Figure 88a,a, thethe slopeslope ofof voltage voltage versus versus current current decreases withwith the the increased increased clamping clamping pressure, pressure, which which means means that the that inner the resistance inner resistance of TEG increased of TEG accordingly.increased accordingly. Combined Combined with the abovewith the presented above presented results, it results, demonstrates it demonstrates that the that thermal the thermal contact resistancecontact resistance of TEG canof TEG be reduced can be reduced for the empty for the air empty gap is air decreased gap is decreased with large with clamping large clamping pressure. Thus,pressure. much Thus, more much inlet more gas heatinlet fromgas heat the from brass the heat brass exchanger heat exchanger can be absorbed can be absorbed by the hot by sides the hot of TEMs,sides of and TEMs, increasing and increasing heat from the heat cold from sides the of TEMscold sides can be of brought TEMs ocanff with be largebrought clamping off with pressure. large Theclamping larger pressure. clamping The pressure larger clamping is, the lower pressure thermal is, the insulator lower thermal is, which insulator contributes is, which to the contributes enhanced outputto the enhanced performance output becaused performance of the becaused lower thermal of the contactlower thermal resistance. contact Therefore, resistance. to ensureTherefore, larger to outputensure larger power output and high power efficiency, and high AETEG efficiency, should AETEG be clamped should as be tight clamped as possible as tight within as possible the allowable within pressurethe allowable of each pressure TEM. of each TEM. 3.6. System Efficiency 3.6 System Efficiency For the 30 TEMs of Bi2Te3 based materials used in TEG, the output of TEG can be expressed For the 30 TEMs of Bi2Te3 based materials used in TEG, the output of TEG can be expressed as as follows [1,18,19]: follows [1,18,19]: X240 VTEG = 240nαPNi∆Ti = n αpi αni)(THi TLi (2) − − VnTnTTTEGi==(-)(-)=1  PNi ipn i i Hi Li (2) i=1 X240 240 RTEG = Ri = 240(nlp/(σpAp)+nln/(σnAn)) (3) RTEG=i=1 RnlAnlA i =240(pppnnn /( )+ /( )) (3) i=1 2 9 α = 22224 + 930.6 0.5 ∆T 0.9905 (0.5 ∆T ) 10− (4) PNi × × i− × × i 2-9× PNi=（ 22224+930.6 0.5TT i -0.9905 （ 0.5  i）） 10 (4) where n is the p-type and the n-type semiconductor galvanic arms number in each TEM and VTEG and RwhereTEG are n is the the open p‐type circuit and voltage the n‐type and internalsemiconductor resistance galvanic of TEG, arms respectively. number inαPNi eachis the TEM relative and V SeebeckTEG and RTEG are the open circuit voltage and internal resistance of TEG, respectively. αPNi is the relative

Energies 2020, 13, 3137 11 of 14

coeﬃcient (V/K), αPi and αni are the Seebeck coeﬃcients of the p-type and the n-type semiconductor galvanic arms, respectively. THi and TLi are the hot side and cold side temperature (K), respectively. lp, 2 σP and Ap are the leg length (m), electricity resistivity (Ωm) and cross-sectional area (m ) of a p-type semiconductor galvanic arm, respectively, while ln, σn and An are the leg length, electricity resistivity and cross-sectional area of an n-type semiconductor galvanic arm, respectively. As shown in Figures5d–8b the output power of the TEG reaches its maximum value (denoted PTEG) when the external load resistance (denoted Rm) is equal to its internal resistance, and is expressed as:

2 ( ) PTEG = UTEG/ 4Rm (5)

The heat input to the hexagonal heat exchanger is obtained as [20]:

Q = G ρ C ∆T = G ρ C (T T ) (6) h h h h h h h h hi − ho where Gh is the volume ﬂow rate of inlet gas, ρh is the density of inlet gas, Ch is the heat capacity of inlet gas at constant pressure, while Thi and Tho are the inlet gas temperature and outlet gas temperature of hot gas, respectively. In this case, the maximum TEG system eﬃciency η can be calculated as follows:

η = PTEG/Qh (7)

Figure9 shows the temperature drop and outlet temperature of hot gas along the brass hexagonal heat exchanger when the inlet gas temperature changes from 155 ◦C to 360 ◦C. On this occasion, the clamping pressure of TEMs above each surface increases to 360 kg/m2, the pumped coolant of ice water mixture (0 ◦C) without radiator is adopted, the ambient temperature is 27 ◦C and the inlet gas backpressure is 70 Pa (the corresponding ﬂow rate is 0.18 m3/h). It can be seen that both the temperature drops and outlet temperature increases with increasing inlet gas temperatures; theEnergies maximum 2020, 13, x temperature FOR PEER REVIEW drop between the inlet and outlet gas is within 50 ◦C. 12 of 15

Figure 9.9.The The decrease decrease in temperaturein temperature and the and outlet the gas outlet temperature gas temperature for different for inlet different gas temperatures. inlet gas temperatures. Experimental results of both the maximum output power and corresponding system efficiency of TEG are shown in Figure 10 with respect to the inlet gas temperatures. As shown in Figure 10, for fixed inlet gas flow rates and backpressures, the maximum output power and system efficiency of TEG based on the coolant of ice water mixture increase with increasing inlet gas temperature. When the maximum inlet gas temperature is 360 ◦C, the outlet gas temperature is 315.1 ◦C, the generated maximum output power of TEG is 11.49 W, and the corresponding system efficiency is 0.96%. For the maximum temperature limitation (360 ◦C) of hot-air blower outlet, the maximum surface temperature of the brass hexagonal heat exchanger shown in Figure6 is only 269.2 ◦C, which is much lower than

(a) (b)

Figure 10. The overall output performance of TEG with respect to the inlet gas temperature. (a) Maximum output power; (b) system efficiency.

Furthermore, to validate the above numerical model of TEG, Figure 11 shows the comparison between the predicted and measured maximum power and corresponding system efficiency with different matched load resistance, the clamping pressure of TEMs above each surface increases is 360 kg/m2, the pumped coolant of ice water mixture (0 °C) without radiator is adopted, the coolant flow is 5000 L/h, the ambient temperature is 27 °C and the inlet gas backpressure is 70 Pa (the corresponding flow rate is 0.18 m3/h). It can be seen that the above numerical model can predict the performances of TEG when the inlet gas temperature changes from 125 °C to 360 °C. At low temperatures, there is a good agreement between the experimental performances and the predicted results. However, at high temperatures, the discrepancy between the experimental performances and predicted values increases evidently and is also acceptable. The main reason is that the heat losses from the hot sides to cold sides of TEMs are not considered in the numerical model, which plays a more important role in the heat transfer at high temperatures. Overall, to estimate the performance and establish the precise model of TEG, the properties of the thermoelectric materials used in TEMs should be assumed to be variables, the heat losses from the uncovered surface, TEMs gap and hot sides to cold sides should be taken into account.

Energies 2020, 13, x FOR PEER REVIEW 12 of 15

Energies 2020, 13, 3137 12 of 14

the operated operation temperature of TEM (330 ◦C). Thus, if the maximum surface temperature of the brass hexagonalFigure 9. heatThe exchangerdecrease in cantemperature be raised and to 330 the◦ C,outlet it can gas be temperature deduced that for the different maximum inlet outputgas power oftemperatures. TEG will approach 20 W, and the corresponding system eﬃciency will be above 1.5%.

(a) (b)

FigureFigure 10. 10.The The overalloverall outputoutput performanceperformance of TEG with respect respect to to the the inlet inlet gas gas temperature. temperature. (a) (a)Maximum Maximum output output power; power; ( (bb)) system system efficiency. eﬃciency.

Furthermore,Furthermore, to to validate validate the the above above numerical numerical model model of of TEG, TEG, Figure Figure 11 11shows shows the the comparison comparison betweenbetween the the predicted predicted and and measured measured maximum maximum power power and and corresponding corresponding system system effi efficiencyciency with with diffdifferenterent matched matched load load resistance, resistance, the the clamping clamping pressurepressure ofof TEMs above above each each surface surface increases increases isis 360 2 360kg/m kg/m2, the, the pumped pumped coolant coolant of of ice ice water water mixture (0(0 ◦°CC)) withoutwithout radiator radiator is is adopted, adopted, the the coolant coolant flow flow is 5000is 5000 L/h, L/h, the ambientthe amb temperatureient temperature is 27 ◦C is and 27 the °C inlet and gas the backpressure inlet gas isbackpressure 70 Pa (the corresponding is 70 Pa (the 3 flowcorresponding rate is 0.18 m flow/h). rate It can is be0.18 seen m3/h). that It the can above be seen numerical that the model above can numerical predict themodel performances can predict of the TEGperformances when the inlet of gasTEG temperature when the inlet changes gas fromtemperature 125 ◦C to changes 360 ◦C. Atfrom low 125 temperatures, °C to 360 °C there. At is low a goodtemperatures, agreement betweenthere is a the good experimental agreement performancesbetween the experimental and the predicted performances results. However, and the predicted at high temperatures,results. However, the discrepancy at high temperatures, between the the experimental discrepancy performances between the and experimental predicted valuesperformances increases and evidentlypredicted and values is also increase acceptable.s evidently The main and reason is also is thatacceptable. the heat The losses main from reason the hot is sidesthat the to cold heat sides losses offrom TEMs the are hot not sides considered to cold insides the of numerical TEMs are model, not considered which plays in the a more numerical important model, role which in the plays heat a transfermore atimportant high temperatures. role in the heat Overall, transfer to estimate at high the temperatures. performance Overall and establish, to estimate the precise the performance model of TEG,and the establish properties the precise of the thermoelectric model of TEG, materials the properties used in of TEMs the thermoelectric should be assumed materials to beused variables, in TEMs theshould heat losses be assumed from the to uncoveredbe variables, surface, the heat TEMs losses gap from and the hot uncovered sides to cold surface, sides TEMs should gap be takenand hot Energies 2020, 13, x FOR PEER REVIEW 13 of 15 intosides account. to cold sides should be taken into account.

FigureFigure 11.11. ComparisonComparison betweenbetween thethe predictedpredicted andand measuredmeasured maximummaximum powerpower andand correspondingcorresponding systemsystem eefficiencyﬃciency of of TEG. TEG.

4. Conclusions Waste heat recovery based on TEMs presents a promising research focus worldwide, but enhancing the output performance and system efficiency of TEG remains a significant challenge. In this study, a TEG system with the low‐temperature and common commercially available Bi2Te3 TEMs and a hexagonal heat exchanger is designed for the waste heat recovery of an industrial hot‐air blower. The influences of different operating conditions such as material, backpressure, inlet gas temperature, clamping pressure and coolant temperature on the temperature distribution, open‐ circuit voltage, the maximum output power and the system efficiency of TEG are reported from the experimental measurements. The experimental results demonstrate that the surface temperature distribution of hexagonal heat exchanger is uniform in each column, and is greatly affected by the adopted material of heat exchanger and the backpressure of inlet gas. The comparisons indicate that the brass hexagonal heat exchanger has better heat conduction, lower backpressure can enhance the gas flow rate and the average surface temperature of heat exchanger (i.e., hot side temperature of TEMs) and the coolant of ice water mixture contributes to lower cold side temperature of TEMs. Furthermore, the maximum output power and system efficiency of TEG is proportional to the practical temperature differences of TEMs caused by inlet gas temperatures, backpressures, clamping pressures and coolant temperature. The designed brass hexagonal heat exchanger has low pressure drop, and it is very suitable for the automotive exhaust thermoelectric generator. Further experiments are planned to optimize the TEG and apply it in the automotive exhaust heat recovery.

Author Contributions: T.L. and Y.Y. performed the experiments and data analysis. R.Q. designed the experiments and wrote the manuscript. Y.C. and B.T. offered valuable discussions in analyses and revised the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding: This research received no external funding.

Acknowledgments: “This work was supported by the National Natural Science Foundation of China (NO. 51977061, 61903129, 51407063)”and “This work was supported by the Doctor Scientific Research Foundation of Hubei University of Technology (NO. BSQD13064)”.

Conflicts of Interest: The authors declare no conflict of interest.

List of Notations

P1 inlet gas temperature sensor T1 inlet gas temperature sensor T2 outlet gas temperature sensor n p‐type and the n‐type semiconductor galvanic arms number

Energies 2020, 13, 3137 13 of 14

4. Conclusions Waste heat recovery based on TEMs presents a promising research focus worldwide, but enhancing the output performance and system efficiency of TEG remains a significant challenge. In this study, a TEG system with the low-temperature and common commercially available Bi2Te3 TEMs and a hexagonal heat exchanger is designed for the waste heat recovery of an industrial hot-air blower. The influences of different operating conditions such as material, backpressure, inlet gas temperature, clamping pressure and coolant temperature on the temperature distribution, open-circuit voltage, the maximum output power and the system efficiency of TEG are reported from the experimental measurements. The experimental results demonstrate that the surface temperature distribution of hexagonal heat exchanger is uniform in each column, and is greatly affected by the adopted material of heat exchanger and the backpressure of inlet gas. The comparisons indicate that the brass hexagonal heat exchanger has better heat conduction, lower backpressure can enhance the gas flow rate and the average surface temperature of heat exchanger (i.e., hot side temperature of TEMs) and the coolant of ice water mixture contributes to lower cold side temperature of TEMs. Furthermore, the maximum output power and system efficiency of TEG is proportional to the practical temperature differences of TEMs caused by inlet gas temperatures, backpressures, clamping pressures and coolant temperature. The designed brass hexagonal heat exchanger has low pressure drop, and it is very suitable for the automotive exhaust thermoelectric generator. Further experiments are planned to optimize the TEG and apply it in the automotive exhaust heat recovery.

Author Contributions: T.L. and Y.Y. performed the experiments and data analysis. R.Q. designed the experiments and wrote the manuscript. Y.C. and B.T. offered valuable discussions in analyses and revised the manuscript. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Acknowledgments: “This work was supported by the National Natural Science Foundation of China (No. 51977061, 61903129, 51407063)”and “This work was supported by the Doctor Scientific Research Foundation of Hubei University of Technology (No. BSQD13064)”. Conflicts of Interest: The authors declare no conflict of interest.

List of Notations

P1 inlet gas temperature sensor T1 inlet gas temperature sensor T2 outlet gas temperature sensor n p-type and the n-type semiconductor galvanic arms number VTEG open circuit voltage of TEG RTEG internal resistance of TEG αPNi relative Seebeck coefficient αpi Seebeck coefficients of the p-type semiconductor galvanic arms αni Seebeck coefficients of the n-type semiconductor galvanic arms THi hot side temperature of TEM TLi cold side temperature of TEM lp leg length of a p-type semiconductor galvanic arm σp electricity resistivity of a p-type semiconductor galvanic arm Ap cross-sectional area of a p-type semiconductor galvanic arm ln leg length of a n-type semiconductor galvanic arm σn electricity resistivity of a n-type semiconductor galvanic arm An cross-sectional area of a p-type semiconductor galvanic arm PTEG maximum output power of TEG Rm external load resistance Qh heat input to the cylindrical heat exchanger Gh volume flow rate of inlet gas ρh density of inlet gas Ch heat capacity of inlet gas at constant pressure Thi inlet gas temperature Tho outlet gas temperature η maximum TEG system efficiency Energies 2020, 13, 3137 14 of 14

References

1. Jaziri, N.; Boughamoura, A.; Müller, J.; Mezghani, B.; Tounsi, F.; Ismail, M. A comprehensive review of thermoelectric generators: Technologies and common applications. Energy Rep. 2019.[CrossRef] 2. Liu, K.; Tang, X.; Liu, Y.; Yuan, Z.; Li, J.; Xu, Z.; Zhang, Z.; Chen, W. High performance and integrated design of thermoelectric generator based on concentric filament architecture. J. Power Sources 2018, 393, 161–168. [CrossRef] 3. Willars-Rodríguez, F.J.; Chávez-Urbiola, E.A.; Vorobiev, P.; Vorobiev, Y.V. Investigation of solar hybrid system with concentrating Fresnel lens, photovoltaic and thermoelectric generators. Int. J. Energy Res. 2017, 41, 377–388. [CrossRef] 4. Demir, M.E.; Dincer, I. Performance assessment of a thermoelectric generator applied to exhaust waste heat recovery. Appl. Therm. Eng. 2017, 120, 694–707. [CrossRef] 5. Meng, F.K.; Chen, L.G.; Feng, Y.L.; Xiong, B. Thermoelectric generator for industrial gas phase waste heat recovery. Energy 2017, 135, 83–90. [CrossRef] 6. Proto, A.; Bibbo, D.; Cerny, M.; Vala, D.; Kasik, V.; Peter, L.; Conforto, S.; Schmid, M.; Penhaker, M. Thermal energy harvesting on the bodily surfaces of arms and legs through a wearable thermo-electric generator. Sensors 2018, 18, 1927. [CrossRef][PubMed] 7. Kim, Y.; Gu, H.M.; Kim, C.; Choi, H.; Lee, G.; Kim, S.; Yi, K.; Lee, S.; Cho, B. High-performance self-powered wireless sensor node driven by a flexible thermoelectric generator. Energy 2018, 162, 526–533. [CrossRef] 8. Leonov, V. Thermoelectric energy harvesting of human body heat for wearable sensors. IEEE Sens. J. 2013, 13, 2284–2291. [CrossRef] 9. Holgate, T.-C.; Bennett, R.; Hammel, T.; Caillat, T.; Keyser, S.; Sievers, B. Increasing the efficiency of the multi-mission radioisotope thermoelectric generator. J. Electron. Mater. 2015, 44, 1814–1821. [CrossRef] 10. Zhang, Y.L.; Cleary, M.; Wang, X.W.; Kempf, N.; Schoensee, L.; Yang, J.; Joshi, G.; Meda, L. High-temperature and high-power-density nanostructured thermoelectric generator for automotive waste heat recovery. Energy Convers. Manag. 2015, 105, 946–950. [CrossRef] 11. Szybist, J.; Davis, S.; Thomas, J.F.; Kaul, B.C. Performance of a Half-Heusler thermoelectric generator for automotive application. SAE Tech. Pap. 2018, 1. [CrossRef] 12. Kim, S.K.; Won, B.C.; Rhi, S.H.; Kim, S.H.; Yoo, J.H.; Jang, J.C. Thermoelectric power generation system for future hybrid vehicles using hot exhaust gas. J. Electron. Mater. 2011, 40, 778–783. [CrossRef] 13. Merkisz, J.; Fuc, P.; Lijewski, P.; Ziolkowski, A.; Galant, M.; Siedlecki, M. Analysis of an increase in the efficiency of a spark ignition engine through the application of an automotive thermoelectric generator. J. Electron. Mater. 2016, 45, 4028–4037. [CrossRef] 14. Fernandez-Yanez, P.; Armas, O.; Capetillo, A.; Martinez-Martinez, S. Thermal analysis of a thermoelectric generator for light-duty diesel engines. Appl. Energy 2018, 226, 690–702. [CrossRef] 15. Fernandez-Yanez, P.; Armas, O.; Kiwan, R.; Stefanopoulou, A.G.; Boehman, A.L. A thermoelectric generator in exhaust systems of spark-ignition and compression-ignition engines. A comparison with an electric turbo-generator. Appl. Energy 2018, 229, 80–87. [CrossRef] 16. Quan, R.; Zhou, W.; Yang, G.Y.; Quan, S.H. A hybrid maximum power point tracking method for automobile exhaust thermoelectric generator. J. Electron. Mater. 2017, 46, 2676–2683. [CrossRef] 17. Quan, R.; Liu, G.Y.; Wang, C.J.; Zhou, W.; Huang, L.; Deng, Y.D. Performance investigation of an exhaust thermoelectric generator for military SUV application. Coatings 2018, 8, 45. [CrossRef] 18. Fraisse, G.; Ramousse, J.; Sgorlon, D.; Goupil, C. Comparison of different modeling approaches for thermoelectric elements. Energy Convers. Manag. 2013, 65, 351–366. [CrossRef] 19. Quan, R.; Wang, C.J.; Wu, F.; Chang, Y.F.; Deng, Y.D. Parameter matching and optimization of an isg mild hybrid powertrain based on an automobile exhaust thermoelectric generator. J. Electron. Mater. 2019.[CrossRef] 20. Niu, X.; Yu, J.L.; Wang, S.Z. Experimental study on low-temperature waste heat thermoelectric generator. J. Power Sources 2009, 188, 621–626. [CrossRef]

© 2020 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/).