《1. Introduction》

1. Introduction

Over a billion people in deprived regions still lack access to electricity, and the situation has been aggravated by the fallout from COVID-19 [1]. The basic provision of electricity to support lighting, communication, and medical care is critically important to promote civilization in deprived areas. Biomass is the primary energy source in deprived regions, and extensive research has explored the generation of electricity by the burning of biomass [2]. The related technology is mainly referred to as biomass-combustionpowered thermoelectric generators (BCP-TEGs). These devices utilize the Seebeck effect to partially convert temperature difference into electricity during biomass combustion [3]. The BCP-TEG is not a new concept, and it has been explored by many researchers in the past two decades since its introduction in 1996 [4]. However, several critical issues remain unclear, including evaluation metrics, efficiency analysis at different levels (i.e., the material, device, and system levels), and a favorable design to obtain a net output power of over 10 W with a portable BCP-TEG. Table 1 presents meaningful comparisons of over two decades’ worth of BCPTEG studies [4–27] by expounding the contributions of previous works, identifying existing issues, and presenting the state of the art in the research field of BCP-TEGs. A comprehensive review of relevant studies reveals that several aspects, including temperature, efficiency, and cooling method, need to be simultaneously controlled in order to develop a standalone, stable, high-capacity, and portable BCP-TEG. The following literature review presents the progress and existing challenges in this field.

《Table 1》

Table 1 Performance comparison of various BCP-TEGs.

CLC: closed-loop cooling; OLC: open-loop cooling; E-heater: electrical heater; G-heater: gas heater; N: number of thermoelectric (TE) modules; m: mass weight of the BCP-TEG; T_h: hot-end temperature; △T: temperature difference; Ptot: total electric power; Pout: electric power output; CHP: combined heat and power; η_HC: heat collection efficiency; η_TE: TE efficiency; η_overall: overall efficiency; η_CHP: combined heat and power efficiency.

— denotes not found or not studied; × denotes not provided.

^a Estimated weight based on water tank volume (L).

^b Estimated results.

Increasing the electric power to 10 W took at least 10 years of effort (with 10.7 W [7] being achieved in 2005 and 12.3 W [9] in 2010) since the first report of the BCP-TEG in 1996 [4]; however, such electric power is not conditioned for a stable output voltage. The augmentation of total electric power (P_tot) remained stagnant during another 10 years until Montecucco’s [21] research in 2017, which reported a new record of 27.0 W. Although the electric power in Goudarzi’s [12] work (155.0 W) is a potential value rather than measured data, the study provides a new idea to improve BCP-TEG performance. Recently, researchers have employed multiple thermoelectric modules (TEMs) to successfully augment the total electric power, reporting 35.0 W [23] in 2018; 75.2 W [25] in 2019; and 12.9 W [24], 62.6 W [26], and 252.0 W [27] in our previous reports between 2018 and 2020. However, a standalone, high-capacity, and portable BCP-TEG that can provide an output power of up to 10 W is still lacking. The mismatch between biomass combustion temperature and the working temperature of the TEM creates considerable difficulties in augmenting the electric power of a BCP-TEG, and the task cannot be accomplished by simply increasing the TEM number. The energy flow of a BCP-TEG is shown in Fig. 1.

《Fig. 1》

Fig. 1. Energy flow of a BCP-TEG indicating various heat losses. q_in: input power; qcomb: total heat flux released by biomass combustion; q_HC: total heat flux collected by heat collector; q_TE: heat flux passing through thermoelectric modules; q_cold: heat flux rejected by the cold ends of BCP-TEG; q_ht: heating power from the radiator; P_tot: total electric power; q₁: incomplete combustion heat loss; q₂: heat loss by flue gases and combustion wall; q₃: heat loss from heat collector; q₄: heat loss by heat sinks, water pumps, and connecting pipes.

Although efficiency is another important issue, only a handful of BCP-TEG studies have provided relevant measurements. Efficiencies are measured at different levels. For example, thermoelectric efficiency (η_TE) measures the conversion ratio of thermal energy passing through a TEM into electric energy, and overall efficiency denotes the conversion ratio of biomass chemical energy into electric power. In other words, overall efficiency covers all possible heat losses (i.e., incomplete combustion heat loss, flue gas heat loss, and convection/radiation heat loss). Therefore, the overall efficiency should be lower than the TE efficiency (η_TE) [28]. Several previous works have reported varying Thermoelectric (TE) efficiency values, including 3.2% in Lertsatitthanakorn’s work [8]; 2.00% in Champier’s work [10]; 1.39% in Goudarzi’s work [12]; 0.65% in Najjar’s work [17]; 5.00% in Montecucco’s work[21]; 1.53% in Obernberger’s work [23]; and 2.84% [24], 3.66% [26], and 2.49% [27] in our previous works. These BCP-TEG studies reported reasonable TE efficiencies because they adopted Bi₂Te₃-based TEMs, which have a maximum TE efficiency of approximately 5%, from various manufacturers around the world. Unlike studies on gas-combustion-powered TEG with an overall efficiency of approximately 3%, previous BCP-TEG studies have yet to report an overall efficiency of up to 1%. Only two previous BCP-TEG studies have reported an overall efficiency of up to 0.5%: 0.63% in Sornek’s work [25] and 0.87% in our previous work [27]. Unfortunately, the abovementioned two BCP-TEGs with overall efficiencies greater than 0.5% have considerable weight, so they are only appropriate for stationary applications.

The main reason for the limited efficiency is the low heat collection efficiency (η_HC), because the flow resistance of the flue gas must be sufficient low; otherwise, the biomass combustion process will be seriously affected. Only three previous works have measured heat collection efficiency, with reported values of 33.3% [12], 23.8% [17], and 34.2% [27]. This status of low heat collection efficiency implies an abundance of opportunities, including augmenting the overall efficiency of BCP-TEGs up to 3% in comparison with existing reports (< 1%) by increasing the heat collection efficiency while ensuring that the biomass combustion is unaffected. In summary, designing an appropriate heat collector with a low flow resistance to augment heat collection efficiency is crucial in developing a high-capacity portable BCP-TEG.

As shown in Table 1, water cooling appears to perform better than air cooling, because the heat capacity of water is much greater than that of air [29]. However, in BCP-TEG studies, the water cooling method is usually open-loop cooling (OLC) from cities’ running water [26] or from a water tank with a large volume, such as 25 L [9], 18 L [10], and 60 L [21]. A BCP-TEG running under water OLC requires plenty of water and limited TEMs. Without these conditions, cold-end heat dissipation causes the water to boil inside the water tank, weakening the performance of the BCP-TEG. Furthermore, water OLC greatly increases the weight of BCP-TEGs, so these types of BCP-TEGs are only appropriate for stationary applications. Therefore, opportunities to develop BCP-TEGs based on water closed-loop cooling (CLC) are abundant. An initial attempt showed that the net electric power output per unit mass weight (P_out/m) in a CLC BCP-TEG reached 1.88 W·kg^–1 , which is much higher than those in previous works [27]. In summary, the inconsistencies of cooling methods lead to difficulties in evaluating various BCP-TEGs. The development of a standalone, high-capacity, and portable BCP-TEG requires the application of the CLC method.  

For a TEM, the long- and short-term working temperatures are quite different. For example, the long-term working temperature of a Bi₂Te₃-based TEM is limited to 523 K, while the corresponding short-term working temperature can be increased to over 573 K. Therefore, the electric power per TEM (Ptot/N) varies greatly. As shown in Table 1, several previous works have reported a considerably large P_tot/N (> 5 W), including 12.30 W [9], 7.38 W [12], 6.00 W [15], 5.00 W [16], 6.00 W [18], and 6.75 W [21]. The working temperature differences of these works fall between 150 and 250 K. It should be noted that a temperature difference of 250 K implies that the hot-end temperature (T_h, K) should be higher than 523 K and that the TEM will experience an aging problem. For other BCP-TEGs with a large P_tot/N, such a property (P_tot/N) may be attributed to the different sizes of TEMs. In general, TEMs come in several sizes, although all are based on Bi₂Te₃; these sizes include 60 mm × 60 mm, 56 mm × 56 mm, 40 mm × 40 mm, and 30 mm × 30 mm. A P_tot/N that is equal to 10 W implies that a heat flux of 200 W should be passed through each TEM; however, doing so is challenging even for a 60 mm × 60 mm TEM, because the heat flux must be as large as 5.6 × 10⁴ W·m^–2 for gas–solid interface heat collection [29]. Therefore, future works on evaluation metrics for BCP-TEGs may consider investigating this issue. In summary, different working temperatures lead to confused output potentials of BCP-TEGs. When developing a stable, highcapacity, and portable BCP-TEG, the working temperature should be maintained within the long-term working temperature range—that is, P_tot/N must be controlled for a particular type of TEM.

Above literature review highlights the need for a high-capacity portable BCP-TEG, and this requirement serves as the motivation of the present work. A 7.6 kg BCP-TEG prototype is illustrated in this work; it provides a favorable result of 23.4 W, which is a noticeable improvement from previous reported values. The temperature distribution, power load characteristics, efficiencies at different levels, and a field test of the proposed method are presented in detail. A comprehensive discussion is also carried out.

The novelty of the present work lies in its pioneering provision of a high-capacity portable BCP-TEG (7.6 kg) that can generate an electric power of 23.4 W. The developed strategy, which interlinks heat collection, heat spread, heat flux matching, and CLC, fills the knowledge gap of an effective method to augment the capacity of a portable BCP-TEG. Moreover, this work is the first to perform a complete set of efficiency analyses for a portable BCP-TEG, including heat collection, TE, overall, and combined heat and power (CHP) efficiencies.

《2. TEG configuration and experimental setup》

2. TEG configuration and experimental setup

《2.1. TEG configuration》

2.1. TEG configuration

Fig. 2 shows a photograph and the structure of the proposed BCP-TEG, which consists of an improved ‘‘rocket-type” biomass combustor, two heat collectors (embedded with a heat spreader), six TEMs, two water heat sinks, and a CHP component. The ‘‘rocket-type” biomass combustor is modified from the convectional ‘‘rocket” stove, which is widely accepted around the world and characterized by good performance in burning biomass [30]. Wood sticks are normally inserted through the bottom opening of a convectional rocket stove [30], making it cumbersome to remove the ash and to introduce a blower to improve the combustion inside the stove. In the present work, the flame holder was moved to a higher location and a fuel-feeding opening was designed above the flame holder, as shown in Fig. 2(c). These modifications isolate the bottom opening for ash removal and completely solve the contradiction of biomass feeding and the intrusion of the heat collector for sufficient heat flux.

《Fig. 2》

Fig. 2. Photograph and structure of the BCP-TEG. (a) Front view; (b) cross-sectional view; (c) TEG configuration; (d) heat collector; (e) dimensions of staggered pin fins.

The CHP component is designed to include a water pump, an expansion vessel, an air blower, a radiator, and an electric energy management system (EEMS). A water pump (HYS-1203B, Shenzhen Hengyuansheng Electronics Co., Ltd., China) and a low-noise air blower (WFB-1212M, Delta Electronic Power Supply Co., Ltd., China) are incorporated in the EEMS. Two direct current–direct current (DC–DC) converters (007915, Shaibang Co. Ltd, China) for electricity conditioning, an electric energy tester (J7-7, JunengXindi Electronic Technology Co., Ltd., China), an indicator of the state of charge (embedded with an alarm), a hot-end temperature indicator, and two output ports (5 and 12 V) are assembled within the EEMS. A battery is externally placed, because a fully charged battery can be taken out for use. The model number of the TEM is TEG1-12708 (Segrea Co. Ltd., China), and its dimensions are 40.0 mm × 40.0 mm × 3.8 mm. One of the contributions of the present work is the development of a novel heat collector (Fig. 2). The heat collector is manufactured from an aluminum block using computer numerical control machining technology, which completely eliminates the possible thermal contact resistance between the heat collector and the heat spreader if they are manufactured separately. Copper rods or aluminum-alloy fins have been used as heat collectors in previous BCP-TEGs [26]. Such heat collectors are subjected to the challenge of an insufficient heat flux for multiple TEMs and the inconvenience of refueling. In addition, gaps between the copper rods and the heat spreader may cause system failure [26]. The dimensions of the staggered pin fins are shown in Fig. 2. The thickness of the heat spreader is 14 mm, which is appropriate for a TEG with six TEMs [28]. The installation pressure of the TEM is 1 MPa [28]. The weight of the whole BCP-TEG, under conditions in which only wood sticks are needed to run it, is 7.6 kg.

《2.2. Experimental setup》

2.2. Experimental setup

Fig. 3 presents the experimental setup. Wood sticks are combusted inside the biomass combustor, and the combustion air is spontaneously induced by the particular configuration of the ‘‘rocket-type” combustor. The flue gases pass through the pin fins, and part of the heat flux from the flue gases is extracted by the heat collector. The extracted heat flux is then distributed by the heat spreader, passing through the TEMs because of the low temperature of the heat sink. To maintain the temperature of the heat sinks, a water pump circulates cooling water between the radiator and the heat sinks, and an air blower delivers heat from the radiator to the atmosphere. Therefore, the heat flux from the radiator is high-quality heating power, which is suitable for heating up a separated room. A total of 17 type-K thermocouples (WREK-101, Shanghai Instrument (Group) Company, China) are installed to measure the hot-end, cold-end, atmosphere, cooling water, and warm air temperatures and to characterize the performance of the BCP-TEG. The accuracy of the thermocouple is 0.5%, and the temperature signals are accumulated with an Agilent 34970A data acquisition (DAQ; 34970A, Agilent Technologies, USA) instrument. The EEMS was developed to stabilize the generated electric power to 12.6 V (for charging the battery) and 5.0 V (universal serial bus (USB) devices). The embedded water pump and air blower are powered by the battery during the startup and become selfpowered by the TEG thereafter. A limited portion of the generated electric power is able to cover the power consumption by the water pump and air blower, and the remaining electric power is charged into the battery or is outputted through the USB ports. When performing a power load test, the EEMS was disconnected, and a 3311F electronic load instrument (Prodigit Technologies, China) was applied to test the electric power under different external loads. The accuracy of the electronic load instrument was found to be 0.2%. The air velocity was measured with a Peakmeter MS6252B turbo-type anemometer (MS6252B, Hyelec, Co., Ltd., China), and the accuracy of the anemometer was found to be 2.0%. A Dali T8 infrared (IR; T8, Zhejiang DALI Technology Co., ltd., China) imager with 25 μm resolution was used to observe the temperature distribution of the entire BCP-TEG. For an uncertainty analysis, interested readers can refer to our previous work [27]; the errors for electric power, heating power, TE efficiency, and overall efficiency were found to be 0.2%, 3.0%, 3.4%, and 3.4%, respectively.

《Fig. 3》

Fig. 3. Experimental setup of the BCP-TEG with various measurements to determine various efficiencies at different levels. T_air: air temperature from radiator; T_wt: cooling water temperature; T_atm: atmosphere temperature; T_h: hot-end temperature; T_h: cold-end temperature; V_air: air velocity from radiator; DAQ: data acquisition.

《2.3. Experimental procedures and cases》

2.3. Experimental procedures and cases

The experimental procedure is divided into two series according to the running conditions. For the power load feature test, the following steps were taken: ① Power up the water pump and air blower using external power sources; ② ignite several wood sticks with several pieces of paper; ③ perform the test when the T_h and cold-end temperatures (T_c, K) are stable; ④ refuel the combustor according to the set point of the hot-end temperature indicator; and ⑤ stop refueling after all measurements are taken, and wait for the burnout of the remaining wood sticks. For the field test, step ① was changed to comprise turning on the EEMS, while the other steps remained the same. During the experiments, the working voltage of the water pump and air blower was kept constant at 11.5 V, and a total of 5.43 W (before the DC–DC converter) was consumed. The electric power consumed by the water pump and air blower varied with time, with 5.43 W as an average value within 5 h in a separated experiment.

A total of 18 experimental cases were carried out. The first 17 experimental cases included measuring the temperature distributions, power load characteristics, and efficiencies at different levels. Hence, the hot-end temperature (T_h1) was maintained at 523 K; then, 17 external loads were explored to reveal the performance of the BCP-TEG. The last experimental case was a field test: burning 1 kg of wood sticks and recording all the necessary parameters to reveal how much electricity could be extracted for outputs. Hence, the EEMS was connected to the TEG, and the water pump and air blower were self-powered up. A separate experiment was conducted to determine the conversion efficiency of the DC–DC converter. A conversion efficiency of 94% is a regular value for a DC–DC converter obtained with the maximum power point tracking algorithm.

Pine wood was used as the fuel for the proposed BCP-TEG. Dry wood sticks were prepared with the dimensions of approximately 20 mm × 20 mm × 300 mm. The lower heat value (LHV) and the density of the dry pine wood were 16.8 MJ·kg^–1 and 482 kg·m^–3 , respectively. The LHV was measured with a calorimeter (5EC5800, Changsha Kaiyuan Co., Ltd., China), and the chemical composition was measured with an elemental analyzer (5E-CHN2000, Changsha Kaiyuan Co., Ltd., China) to be C = 41.89%, H = 4.51%, O = 40.16%, N = 0.22%, and S = 0.05%. No air blower was used to accelerate the wood combustion inside the BCP-TEG, as the rising flow of flue gases is a typical natural ventilation phenomenon.

《2.4. Data evaluation》

2.4. Data evaluation

The biomass energy is the input power of the whole BCP-TEG, which is defined as follows: 

where q_in is the input power (W), and m_wd is the mass weight of wood sticks (kg; m_wd = 1 kg) consumed during △t (s) and LHV is the lower heat value of the wood sticks. Not all biomass energy can be released, due to incomplete combustion. The result is the incomplete combustion heat loss (q₁, W). A substantial heat flux is released through combustion; part of the heat flux is absorbed by the heat collector (q_HC, W), while the remainder is the heat loss by flue gases and combustion walls (q₂, W) is lost by the flue gases and combustion walls. Unfortunately, it is difficult to measure q₁ and q₂. As mentioned previously, not all of the heat flux absorbed by the heat collector can be forced to pass through the TEMs; other heat dissipation is the heat loss from heat collector (q₃, W) from the side surfaces of the heat collector through convection and thermal radiation. q₃ is estimated as follows:

where A_HC is the heat transfer area of heat collector (m² ), THC is the outside surface temperature of heat collection (K), T_atm is the atmosphere temperature (K),  is the Stefan–Boltzmann constant (W·m^–2 ·K^–4 ), the convective heat transfer coefficient (h_air, W·m^–2 - ·K^–1 ) and emissivity (e) are introduced in detail in our previous study [28]. Part of the heat flux passing through the TEMs is converted into total electric power (P_tot, W); the rest (q_cold) is rejected by the heat sinks. q_cold is divided into two parts: heat loss by heat sinks, water pumps, and connecting pipes (q₄, W) through natural convection and thermal radiation (from the heat sink surfaces, water pumps, and connecting pipes), and heating power from the radiator (q_ht, W). q₄ and q_ht are calculated as follows [28]: 

where A_PH is the outside surface area of heat sinks, pipes, and water pumps (m² ), T_PH is the surface temperature of heat sinks, pipes, and water pumps (K), c_p is the heat capacity of air (kJ·kg^–1 ·K^–1 ), ρ_air is the air density (kg·m^–3 ), V_air is the air velocity from radiator (m·s^–1 ), A_R is the area of radiator (m² ), T_air is the air temperature from radiator (K).

As a result, the heat flux collected by q_HC and the heat flux passing through the TEMs (q_TE, W) can be derived as follows:

The heating power from the radiator (q_ht) can be used to heat up a separate space. Therefore, the present BCP-TEG is a CHP solution that provides clean heat and electricity simultaneously. Various efficiencies at different levels can be derived. The heat collection efficiency is a critical parameter and is calculated as follows:

The TE, overall, and CHP efficiencies represent the BCP-TEG performance at different levels. The TE efficiency measures the TEM performance at the device level, while the overall efficiency describes the BCP-TEG performance at the system level. Furthermore, the CHP efficiency is a reflection of how high in quality the energy conversion is in terms of energy utilization. The TE, overall, and CHP efficiencies are calculated as follows:

《3. Results and discussion》

3. Results and discussion

《3.1. Temperature distribution》

3.1. Temperature distribution

Fig. 4 shows a thermal image of the BCP-TEG and the distributions of the T_h and T_c along the heat collector. The thermal imager was calibrated in a separate experiment with thermocouples, and the emissivity of the aluminum surface was set at 0.4 [28]. The thermal image indicates that the wall temperature of the biomass combustor without insulation is quite high, revealing that insulation could be an effective way to increase the flue gas temperature. The details of the temperature measurements are presented in Fig. 4(b). The T_h is unevenly distributed, indicating that not all the TEMs are operating at T_h = 523 K. Consequently, the potentials of some TEMs are not fully utilized because of the compromise made to protect the TEMs running at high hot-end temperatures. Therefore, the hot-end temperature varies between 503 and 523 K, indicating a 20 K difference along the heat collector. The thickness of the heat spreader (14 mm in the present work) is an essential parameter that helps to transfer excessive heat flux from areas with high heat flux to areas with low heat flux. In cases with a thinner heat spreader, the temperature unevenness deteriorates further. A thickness of 14 mm is an optimized value for a heat spreader fitting three TEMs [28]. The T_c undergoes a slight variation between 336 and 339 K and is controlled with the water CLC method to reach the minimum power consumption and mass weight of the CHP component. Hence, the working temperature difference of the TEMs is between 164 and 188 K.

《Fig. 4》

Fig. 4. (a) Thermal image of the BCP-TEG; (b) distributions of T_h and T_c.

《3.2. Power generation》

3.2. Power generation

A power load feature test is used to determine the optimized external load and maximum electric power, as shown in Fig. 5. The generated voltage increases with the external load resistance, whereas the load current has the opposite trend. The maximum electric power was found to be 23.4 W at load resistance R_ld = 14 Ω. In addition, the electric power first increased with the external load, reaching a maximum value, and then underwent a decreasing trend when the external load was further increased. Compared with the previous works listed in Table 1, the present work is the first to report a portable BCP-TEG with a total electric power over 20 W, which is greater than those of other BCP-TEGs with comparable mass weights (5.0 W [20] and 12.9 W [24]). The underlying measure to reach this result is the combination of a compact CLC method, a well-designed heat collector, and a ‘‘rocket-type” biomass combustor. Properly distributed pin fins ensure a low flow resistance for the flue gas and sufficient heat flux for power generation. Furthermore, the heat spreader having sufficient thickness ensures a relatively even hot-end temperature, which promotes the extraction of electric power from each TEM.

《Fig. 5》

Fig. 5. The power load features of BCP-TEG when the hot-end temperature is maintained at 523 K. R_ld: load resistance.

《3.3. Efficiency》

3.3. Efficiency

Various heat fluxes can be calculated using Eqs. (1–6). The q_in, q₃, q₄, q_ht, q_HC, q_TE are 2393.2, 49.8, 42.0, 750.4, 865.6, and 815.8 W, respectively. In this way, the energy flow distribution of the BCP-TEG was obtained and is shown in Fig. 6. The incomplete heat loss and the heat losses by flue gases and combustion walls are 1527.6 W, which is quite high and is applicable for cooking. Separate experiments were performed, and the results show that the proposed BCP-TEG is applicable for cooking, and that the cooking function has no influence on the CHP cogeneration. The q_ht can be used to heat up a separate space. Therefore, the proposed BCPTEG is a CHP solution that simultaneously provides clean heat and electricity.

《Fig. 6》

Fig. 6. Energy flow distribution of the BCP-TEG when the hot-end temperature (T_h1) is maintained at 523 K. HC: heat collector.

Various efficiencies at different levels can be derived based on Fig. 6 and Eqs. (7–10). The η_HC, η_TE, η_overall, and η_CHP are 36.2%, 2.87%, 0.98%, and 32.3%, respectively. The η_overall is extremely low (less than 1%), even though the η_TE is near 3%. This result reveals that a large portion of the heat released by biomass combustion is lost by flue gases. Recovering this heat loss from flue gases is an interesting problem; however, we chose to abandon this heat loss in the present work because compactness and a low weight were essential concerns. There are several possible methods to collect extra heat for CHP applications. For example, water jackets can be used as combustion walls, and finned heat exchangers can be placed at the combustor outlet. The η_overall reported in the present work is greater than those in previous studies (0.63% in Ref. [25] and 0.87% in Ref. [27]). Future investigations could focus on augmenting the heat collection efficiency. The η_CHP is as high as 32.3%, which indicates that the present BCP-TEG shows potential for clean heating and electricity. 

Comparisons of η_TE and heat collection performance between the present results and those in previous studies, as shown in Table 1, reveal the existing patterns in the research field of BCPTEG. ① Increasing the electric power requires the incorporation of multiple TEMs, but the inherent uneven T_h distribution restricts the performance of a certain portion of the TEMs. This restriction, caused by the mismatch of working temperature, reduces the η_TE. ② A low pressure drop caused by the installation of the heat collector is required for the natural combustion of biomass. Otherwise, a complex controlling system for forced combustion is needed, and more electric power must be consumed. Consequently, the efficiency of the heat collection under the conditions of a low pressure drop and low flue gas velocity is limited. ③ The goverall is the product of η_HC, η_TE, and η_comb, and η_comb is the combustion efficiency. Hence, the overall efficiency of BCP-TEGs is limited by the abovementioned two patterns.

《3.4. Field test》

3.4. Field test

The purpose of the power load feature test described in Section 3.2 was to determine the maximum electric power of the present BCP-TEG. However, the generated electric power cannot be fully extracted because of the mismatch between the internal electrical resistance and the external load resistance in real applications. Hence, a field test was carried out to determine how much electric energy could be outputted by burning 1 kg of wood sticks in a manner consistent with that described in Sections 3.1–3.3— that is, with the T_h controlled at T_h1 = 523 K. The EEMS was connected, making the BCP-TEG run in a standalone mode. It should be noted that the battery used to store electric energy had to fully accept all outputted electric power from the BCP-TEG and could not enter a saturated charging state. When the battery shifted to the saturated charging state, it had to be replaced with a new, uncharged one. Fig. 7 presents the BCP-TEG performance when burning 1 kg of wood sticks. The T_h is near 523 K, and the output voltage (U) varies between 12.0 and 13.5 V (before the DC–DC converter), whereas the voltage after the DC–DC converter is set at 12.6 V. The voltage before and after the DC–DC conversion could not be maintained at a constant value when the battery fully accepted all the generated electric power. However, the voltage after the DC–DC conversion stabilized at 12.6 V when the battery underwent a saturated charging state, and the voltage before the DC–DC conversion rose to a high value, indicating that the generated electric power was not fully extracted. When the output voltage was greater than 13.5 V (e.g., at 80 min in Fig. 7(a)), the battery was near the saturated charging state and was replaced with a new uncharged battery.

《Fig. 7》

Fig. 7. Performance of the BCP-TEG when burning 1 kg of wood sticks. (a) Time evolutions of voltage (U), current (I), and T_h; (b) time evolutions (t) of total electric power (P) and electric energy (E).

The measured electric power was not as great as the theoretical electric power calculated in Section 3.2, because the electrical resistance of the battery does not match the internal resistance of the BCP-TEG. The total electric energy generated by burning a kilogram of pine sticks was recorded with an electric energy tester. The results showed that a total of 35.2 W·h of electric energy was generated, and that 10.6 W·h (5.43 W × 1.95 h) was consumed by the water pump and air blower. The average electric power consumption by the water pump and air blower was introduced in Section 2.3. Hence, 23.1 W·h—that is, (35.2–10.6) W·h × 94%— was stored inside the batteries and could be used to fully charge a 3.7 V battery of 6.2 A·h—that is, 23.1 W·h/3.7 V = 6.2 A·h. This analysis reveals the possibility of fully charging a smart phone by burning a kilogram of wood sticks with the present BCP-TEG (e.g., iPhone 12 Pro with a battery of 3.69 A·h). This result is because the biomass energy of a kilogram of wood sticks is much greater than the energy stored inside a smart phone, even with an overall efficiency of less than 1%. Furthermore, this work is the first to report the quantity of electricity that can be outputted with a BCP-TEG by burning a kilogram of wood sticks. A previous study showed that a minimum electric energy of 3 W·h per day is required in deprived regions [31]. Therefore, the present BCP-TEG meets the electricity demand in deprived regions and can be extended to other possible applications, such as pumping water or powering up small medical devices.

《3.5. Discussion》

3.5. Discussion

3.5.1. The importance of CLC

A critical parameter for BCP-TEGs is net power density (P_out/m), which measures the compactness of power supply sources on the basis of mass weight. Fig. 8 shows a comparison of the net power densities of various BCP-TEG studies. One of our previous works [26] achieved the highest net power density of 4.79 W·kg^–1 . However, the BCP-TEG reported in that work [26] needs to operate using running water from rivers or lakes. In other words, sufficient water must be supplied, and the cooling water from the heat sinks is abandoned. Therefore, the cooling method in Ref. [26] is based on the OLC method. Other BCP-TEG studies based on the OLC or air-cooling methods have low net power densities (0.27 W·kg^–1 in Ref. [7], 0.49 W·kg^–1 in Ref. [9], 0.28 W·kg^–1 in Ref. [10], 0.25 W·kg^–1 in Ref. [11], 0.32 W·kg^–1 in Ref. [21], and 0.89 W·kg^–1 in Ref. [24]). The underlying reasons include the poor cooling efficiency of the air-cooling method and the large water tanks incorporated in the OLC method. Furthermore, the water temperature gradually increases after the startup of a BCP-TEG based on the OLC method, because no radiators are employed to dissipate heat from the water. To augment the net power density and ensure that the BCP-TEG is able to function everywhere, the CLC method should be developed—that is, extra radiators and air blowers should be applied to dissipate heat from the cooling water into the atmosphere. However, the CLC method involves a number of components (i.e., an expansion vessel, water pump, radiator, and air blower) and consumes electric power. Therefore, the ultimate performance of BCP-TEGs based on the CLC method requires careful examination. One of our previous works [27] successfully applied the CLC method in a medium-sized BCP-TEG and achieved promising results (1.88 W·kg^–1 ). In the present work, the CLC method holds potential even for a portable BCP-TEG, regardless of extra components that need to be installed to dissipate heat. The net power density reaches 2.41 W·kg^–1 and can still be improved further, because stainless steel plates with a thickness of 2 mm are used in the present work and these can be substantially reduced. This discussion reveals the importance of the CLC method in developing standalone, high-capacity, and portable BCP-TEGs. The underlying mechanism is the high thermal capacity of water to evenly cool down the cold end of the TEG and the dissipation of heat from the water using light, compact, and wellfinned aluminum radiators and plastic air blowers. The ultimate electric power output compensates for the power consumption of the water pumps and air blowers. In fact, the proposed BCP-TEG generates more electric power than BCP-TEGs based on air cooling or the OLC method with large water tanks. 

《Fig. 8》

Fig. 8. Comparison of the net power densities of various BCP-TEGs with different cooling methods.

《Fig. 9》

Fig. 9. Comparison of the η_overall of various BCP-TEGs.

3.5.2. Difficulty of reaching an overall efficiency of 1% for BCP-TEGs

Low efficiency is the major disadvantage of TEGs, especially BCP-TEGs. For gas-combustion-powered TEGs, the overall efficiency can reach 3% for direct gas-combustion-powered TEGs [32] or 2.5% for catalytic gas-combustion-powered TEGs [33]. A comprehensive review of gas-combustion-powered TEG has been provided by Mustafa et al. [34].

Relatively few studies have reported on the η_overall of BCP-TEGs, as shown in Table 1 and Fig. 9. The η_overall of BCP-TEGs is still below 1% after two decades of investigations. An in-depth analysis indicates that the η_HC of BCP-TEGs is much lower than those (larger than 80%) of gas-combustion-powered TEGs. Hence, overall efficiency is seriously affected. Only three BCP-TEG studies have reported heat collection efficiencies (33.3% in Ref. [12], 23.8% in Ref. [17], and 34.2% in Ref. [27]). As shown in Fig. 9, the overall efficiency of the present BCP-TEG is 0.98%, which is higher than those in previous reports. The reason behind the low η_HC is the difficulty in collecting heat from the flue gases of biomass combustion, which have substantial soot and a low flowing velocity. The low flowing velocity implies that biomass combustion is sensitive to flowing resistance, but installing a heat collector with sufficient finned area to collect adequate heat flux will cause substantial flow resistance. Consequently, a compromise between η_HC and good biomass combustion needs to be made, which will further affect the η_overall. A possible solution is to completely control the biomass combustion process—that is, for all combustion air to be provided by an air blower. This solution for simultaneously enhancing the combustion process and heat collection requires experimental verification in future work.

3.5.3. Problem of heat flux per TEM

Optimizing the heat collection is an important aspect of augmenting the capacity of BCP-TEGs. However, a well-functioning heat collector needs a well-designed heat spreader—that is, the heat flux should evenly pass through the TEMs. The T_h must be evenly distributed. In the present work, a heat spreader with a thickness of 14 mm was designed to spread the heat flux and thereby ensure the compactness and high capacity of the proposed BCP-TEG. Therefore, the heat flux per TEM should be investigated and compared with those in previous works. However, a limited number of previous works have measured η_TE. Therefore, only those studies that reported P_tot/N > 3 W and η_TE (or T_h/T_c temperatures) are used for further discussion. For the studies that do not report the η_TE, the T_h and T_c are used to theoretically predict their TE efficiencies. It should be noted that the deduced η_TE are then used to calculate the heat flux per TEM (Q_TEM) in Fig. 10, and certain errors can exist. Nevertheless, Fig. 10 represents a well-estimated Q_TEM. The predicted TE efficiency can be theoretically predicted as follows [35]:

The figure of merit (ZT_h) is assumed to be 0.8. In Eq. (11), r, w, n, and L are the thermal contact ratio, ratio of ceramic thickness to the length of the TE leg, electrical resistivity ratio, and length of the TE leg, respectively. The η_TE,theo vary between 1.36% and 4.9%, depending on the T_h and T_c. As shown in Fig. 10, the heat flux per TEM varies. An important reason is the different sizes of the TEMs; for example, HZ-20, which is used in several works [4,6,7,9], measures 75 mm × 75 mm. Therefore, HZ-20 allows a high heat flux of up to 500 W, according to its datasheet. However, BCP-TEGs using HZ-20 TEMs work at about 50% of the permitted Q_TEM. Consequently, the potential of the TEM is not totally released. In Goudarzi et al.’s [12] work, however, Q_TEM is higher than 500 W, even though the TEM adopted measures 56 mm × 56 mm. The TEM is thus overused, according to the datasheet of TEP1-12556-0.6, as the permitted Q_TEM for TEP1-12556-0.6 is less than 300 W. Recent studies have begun to balance the heat flux distribution and TEM area; in other words, research tends to employ TEMs with dimensions of 40 mm × 40 mm, which work with a permitted heat flux of approximately 150 W. This facilitates heat flux management and helps to release the potential of the TEM. Fig. 10 shows that performing heat flux matching design is crucial (e.g., Ranman et al.’s work [14]). The present Q_TEM is 136 W, which is close to the permitted Q_TEM (144 W) and indicates that the potential of the TEM has been released. 

《Fig. 10》

Fig. 10. Comparisons of heat flux per TEM in various BCP-TEGs.

《4. Conclusions》

4. Conclusions

A compact, standalone, high-capacity, and portable BCP-TEG was designed and tested in this work. Detailed experiments were performed to investigate the temperature distribution, power load characteristics, and various efficiencies at different levels. A field test was also conducted. The following conclusions can be drawn based on an analysis of the results and a thoughtful discussion.

(1) The developed 7.6 kg BCP-TEG can cogenerate an electric power of 23.4 W and a clean heating power of 750 W, with a net power density of 2.41 W·kg^–1 . A 3.7 V battery of 6.2 A·h can be fully charged by burning 1 kg of wood sticks. The overall and TE efficiencies are 0.98% and 2.87%, respectively. The overall efficiency is higher than those in previous reports.

(2) A detailed analysis indicates that the CLC method can be used to augment the capacity of BCP-TEGs, although several other components (i.e., an expansion vessel, water pump, radiator, and air blower) need to be added and extra power is consumed. The ultimate net power density (2.41 W·kg^–1 ) is higher than those based on the OLC method and incorporating large water tanks.

(3) The balance between the heat collection and flow resistance of flue gas is essential in improving the BCP-TEG performance. The heat collection efficiency should be as high as possible when the biomass combustion is well sustained. In the future, completely controlling the air for biomass combustion using an air blower could be an interesting solution to increase the overall efficiency.

(4) A heat flux matching design should be considered first when developing a high-capacity BCP-TEG. The potential of each TEM can be released only when the heat flux matching, large heat collection efficiency, and heat flux spreading are simultaneously fulfilled.

《Acknowledgments》

Acknowledgments

This work is supported by the key program of the Natural Science Foundation of Zhejiang Province (LZ21E060001) and the key R&D plan of Zhejiang Province (2020C03115).

《Compliance with ethics guidelines》

Compliance with ethics guidelines

Guoneng Li, Jie Ying, Minbo Yi, Youqu Zheng, Yuanjun Tang, and Wenwen Guo declare that they have no conflict of interest or financial conflicts to disclose.