Performance assessment and optimization of a heat pipe thermal management system for fast charging lithium ion battery packs

International Journal of Heat and Mass Transfer 92 (2016) 893–903

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer journal homepage: www.elsevier.com/locate/ijhmt

Performance assessment and optimization of a heat pipe thermal management system for fast charging lithium ion battery packs Yonghuang Ye a, Yixiang Shi b, Lip Huat Saw a,c, Andrew A.O. Tay a,⇑ a

Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117576, Singapore Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing 100084, China c Department of Mechanical and Materials Engineering, University Tunku Abdul Rahman, Kajang 43000, Malaysia b

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 19 May 2015 Received in revised form 4 September 2015 Accepted 17 September 2015

Thermal management system is critical for the electric vehicles and hybrid electric vehicles. This is due to the narrow operating temperature range for lithium ion batteries to achieve a good balance between performance and life. In this study, heat pipes are incorporated into a thermal management system for prismatic or pouch cells. Design optimizations focusing on increasing the cooling capacity and improving temperature uniformity of the system are undertaken through sensitivity studies. Subsequently, empirical study is carried out to assess the thermal performance of the optimized design integrated with prismatic cells at the unit level and the battery pack level. The results confirm that the optimized heat pipe thermal management system is feasible and effective for fast charging lithium ion battery packs. A delay quench cooling strategy is also proposed to enhance the performance of the thermal management system. Ó 2015 Elsevier Ltd. All rights reserved.

Keywords: Lithium ion battery Thermal management Heat pipe Electric vehicle

1. Introduction The lithium-ion battery is regarded as an optimum energy storage device for electric vehicles (EVs) and hybrid electric vehicles (HEVs) [1,2]. Its thermal management, however, is still one of the key issues restricting the application of lithium-ion batteries on EVs and HEVs. Various studies have concluded that the high temperature accelerates the capacity degradation and shortens the battery life [3,4]. Thermal management systems (TMSs) are therefore crucial for EVs and HEVs to control the operating temperature of batteries within an appropriate range. Besides, a TMS is also essential for preventing batteries from thermal run away and catching fire [5]. It has been elucidated by various studies that the appropriate operating temperature range for lithium-ion batteries is 25–40 °C, within which the lithium-ion battery achieves a good balance between performance and life [6,7]. For batteries operated under moderate conditions, many researchers adopted a maximum surface temperature limit of 50 °C when designing the battery TMS [8,9]. It is also desired to maintain the temperature difference within a module to be below 5 °C to avoid large cell-to-cell imbalance which may accelerate the degradation of batteries [6]. ⇑ Corresponding author. Tel.: +65 65162207; fax: +65 67791459. E-mail addresses: (A.A.O. Tay).

[email protected]

(Y.

Ye),

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2015.09.052 0017-9310/Ó 2015 Elsevier Ltd. All rights reserved.

[email protected]

Besides, due to the limited space in an EV, a TMS is also required to be compact and lightweight [10]. There have been various types of TMSs developed for EVs. To date, most of the studies on TMSs are focused on the development of active cooling systems in which the coolant is air, oil, or water-ethylene glycol [1,11–16]. However, air cooling may not be sufficient if the battery module/pack is under stressful operating conditions (e.g. fast charging) or in a thermal abuse condition [17,18]. When using liquid as the coolant, preventing liquid leakage is challenging and costly. The use of dielectric fluids will also increase the cost of the system. Furthermore, the relatively high pressure drop across the liquid-cooled heat exchangers will lead to significant additional energy consumption [1]. On the other hand, the phase change material (PCM) TMS performs better in terms of battery-pack compactness [19] and temperature uniformity [20–22]. However, the PCMs alone are still insufficient for high heat fluxes due to the low thermal conductivity of the PCM. Another relatively novel concept of a TMS is the heat pipe thermal management system (HPTMS). Heat pipes have been widely used in many industrial applications [23–27], and they are known as thermal superconductors operating on the principle of high heat transfer through evaporation [28]. The effective thermal conductivity of heat pipes can reach up to 90 times higher than that of a copper bar of the same size [29]. For heat pipes with a metal sintered powder wick, the heat transfer rate in the radial direction through liquid evaporation is much greater than the heat transfer

894

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

Nomenclature Symbols A Cp ki Q Qeff Tf Ti Tin Tj Tout Tw

battery area subjected to cooling (m2) specific heat capacity of the battery (J kg1 K1) thermal conductivity (i = x, y, z) (W m1 K1) transient heat generation rate (W) effective heat transfer rate gained by the cooling water (W) coolant temperature (°C) condenser temperature (i = 1, 2, 3, 4) (°C) inlet temperature of the coolant (°C) battery surface temperature (j = a, b, c) (°C) outlet temperature of the coolant (°C) measured average temperature of the HPCP (°C)

rate along the heat pipe envelope. As a result, an isothermal temperature profile can be achieved along the evaporator section of the heat pipe [30], and therefore ensures a good uniformity of the battery temperature when heat pipes are applied to the thermal management of batteries. Wu et al. [31] attached two heat pipes with aluminum fins to the battery wall to improve the heat transfer process. Their simulation results showed that the application of the heat pipe can significantly reduce the temperature rise. Rao et al. [9] developed a HPTMS for prismatic cells, and the experimental results showed that the maximum temperature can be controlled below 50 °C when the heat generation rate of the cell was lower than 50 W. A similar design utilizing an aluminum cooling plate embedded with heat pipes was numerically investigated by Greco et al. [32]. Wang et al. [33] proposed a HPTMS design for cooling and heating purposes, the system is able to control the battery temperature below 40 °C if the heat generation is less than 10 W/cell. Most of these TMSs are focused on low C_rates (i.e. less than 1 C [31]; 2 C [33]; 3 C [34]) operating conditions where the heat generation in the cell is relatively small. A C_rate is a measure of the rate at which a cell is charged/discharged relative to its maximum capacity. A 1 C_rate current will discharge the entire capacity of the cell in 1 h, and an n C_rate current is n times that of a 1 C_rate current [35]. Besides, the optimization of these HPTMSs has not been attempted and most of the systems have not been experimentally validated with integration of real cells. In fast-charging applications, the battery cell needs to be charged at high C_rate to achieve full capacity in 10 min [36]. A large amount of heat is generated which will cause a rapid temperature rise of the cell [37]. Besides, under some severe conditions, such as sharp acceleration, over-discharge of cell, cell internal short circuit, etc., the heat generation within cells would be significantly increased. This may possibly lead to excessive temperature throughout a module/pack or thermal runaway of cells. Hence, a reliable thermal management system must be designed to cater for these severe operating conditions and maintain the cell within the optimum operation temperature limits. Existing TMSs utilizing heat pipes are not designed to dissipate a large amount of heat generation from the battery pack. The heat pipe may reach its dry-out point when the heating power at the evaporator section exceeds a critical value [38]. In addition, due to the poor thermal conductivity of the active material layer and separator in the battery, the internal temperature of the cell is higher than the skin temperature of the cell [37]. Hence, a lower limit of the maximum surface temperature of 40 °C is a safer criterion than 50 °C when designing a TMS. In view of the above, the aim of this work is two fold: first, to design and optimize a HPTMS, so that the thermal management system can meet the requirements for fast charging and extreme

Greek letters q density of the battery (kg m3) Subscripts, superscripts and acronyms EV electric vehicle HEV hybrid electric vehicle HPCP heat pipe cooling plate HPTMS heat pipe thermal management system PCM phase change material TMS thermal management system

operating conditions. In order to meet this target, sensitivity studies on various design parameters, including the working fluid of heat pipes, flow rate, coolant temperature, operating orientation, the number of heat pipes in a HPCP and with/without cooling fins, will be carried out to identify the influence of each factor on the thermal response of the HPTMS. Second, the effectiveness of the system will be assessed with actual prismatic cells, and a numerical model was developed to predict the thermal performance of the HPTMS. Finally, the effectiveness of different cooling approaches, viz. ‘‘constant 25 °C”, ‘‘constant 15 °C”, ‘‘decreasing”, ‘‘quench”, and ‘‘delay quench” cooling, in controlling the cell temperature were investigated.

2. Conceptual design of a HPTMS Fig. 1(a) shows a battery pack design, in which there are 110 prismatic cells (140
65
15 mm, 10 Ah) in a 10
11 arrangement. In the design, cells and heat pipe cold plates (HPCPs) are stacked alternately. That is, each HPCP is sandwiched between two cells and each cell is sandwiched between two HPCPs. HPCPs are in contact with both sides of each cell, a layer of thermal grease is applied at the interfaces between the cells and the HPCPs, such that good thermal contact can be ensured. The HPTMS is designed for the thermal management of the battery pack during 8 C_rate of fast-charging. During the operation, the heat generated within the cells is conducted through the evaporator sections of the heat pipes to the condenser sections, and from there to the circulating coolant over the condenser sections. A typical unit of the HPCP is shown in Fig. 1(b). Each of the HPCP comprises a number (e.g. 4) of heat pipes and two copper-plate heat spreaders of size matching the battery cell side surface. The evaporator sections of the heat pipes are flattened, and the two copper plates are soldered to the

Fig. 1. Schematic of the conceptual design of the HPTMS for a battery pack.

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

flattened surfaces of the heat pipe. Cooling fins are push-fitted and soldered to the condenser sections of the heat pipes so as to enhance the cooling performance. Making use of the super high thermal conductivity of the heat pipe, an isothermal temperature profile can be achieved along the evaporator [30]. Due to symmetry of the battery pack, a HPCP is only required to dissipate an amount of heat that equal to the total heat generation in one battery as both sides of the battery are subjected to cooling in a symmetric arrangement. A safety factor of 2.0 is adopted here in order to secure the safety of the battery pack in the worst case scenario. The experimental measurement results using an adiabatic calorimeter (Accelerated rate calorimeter, THT., Ltd) plotted in Fig. 2, which has been presented in our earlier study [39], showed that the averaged heat generation rates within a cell at 3 C_rate, 5 C_rate and 8 C_rate charging are 10.5 W, 25.4 W and 54.4 W, respectively. For the battery cells in this study, the 1 C_rate corresponds to a 10 A current which fully discharges the cell in 1 h. Since the HPTMS is designed for thermal management of batteries during 8 C_rate of charging, the optimized HPCP is required to have a cooling capacity of 108.8 W. In order to deliver the optimum performance of the battery, the HPTMS is also required to control the temperature of the battery within the range of 25 °C–40 °C and control the temperature difference below 5 °C [6,7]. Various steps in optimizing the design of the HPTMS are: 1. Selecting appropriate single heat pipes with high maximum heat transfer limits and low thermal resistances (diameter, working fluid, flattened thickness and the number of heat pipes). 2. Choosing appropriate operating conditions (flow rate, coolant temperature and operating orientation). 3. Optimizing the design by soldering copper heat spreaders to the heat pipe evaporator sections to improve the temperature uniformity of the battery surface, and adding cooling fins at the condenser sections of heat pipe to enhance the heat dissipation performance. To optimize the design, sensitivity studies on various design parameters, including the working fluid of heat pipes, flow rate, coolant temperature, operating orientation, the number of heat pipes in a HPCP and with/without cooling fins, were undertaken to determine the most influential factors on the thermal performance of the HPTMS.

Fig. 2. Experimental measurement of heat generation rate of a cell during charging at different C_rates.

895

3. Experimental parametric study on the performance of a HPCP In order to provide better heat transfer between the flat battery cell surfaces and the heat pipes, a heat pipe cold plate (HPCP) is used instead of individual heat pipes. An initial study on a baseline HPCP was conducted as follows. The evaporator sections of 4 pieces of 8.0 mm diameter sintered copper powder heat pipes (with water as the working fluid) were flattened to 3.0 mm in thickness. Two pieces of copper plate heat spreaders (0.5 mm in thickness) matching the size of one prismatic battery were soldered onto the flattened evaporator sections of the heat pipes. As mentioned, parametric studies will be carried out before the HPCP is integrated into the HPTMS for the thermal management of batteries. The averaged heat generation rate of a cell has been used to determine the required cooling capacity of the HPCP under worst case scenario in Section 2. Since the targeted cooling capacity of the HPCP is higher than the actual heat generation rate of the cell, plated heaters matching the side-surface of a battery cell were used to simulate the heat load from cells. The constant heat loads supplied by the plate heaters also help to ease the performance assessment of the HPCP in the parametric study. A schematic of the cooling loop of the experimental setup is shown in Fig. 3. In the cooling loop, a thermostatic water bath comprising a heating component and a refrigerant system is used to regulate the coolant temperature in the cooling loop, a gate valve is used to cut off the coolant flow, and a ball valve is used to regulate the coolant flow rate. The details of the test section illustrated in Fig. 3 to test the performance of the HPCP are shown in Fig. 4. In the HPCP shown in Fig. 4(b), the 4 heat pipes are distributed evenly with a constant spacing of 16 mm so as to secure good temperature uniformity. In the test section, a plate heater was attached to each side of the HPCP to simulate the heat load from the cells. The adiabatic sections of the heat pipes were insulated by a Teflon block. The bare condenser sections of the heat pipes in the HPCP were cooled by the liquid cooling loop as shown in Fig. 3. Thermal grease was applied to the gap between the plate heaters and the surface of the HPCP to ensure good thermal contact and uniform heat supply. T-type thermocouples were calibrated in a temperature calibrator (FAST-CAL, Isothermal Technology Limited) to ±0.1 °C accuracy, and were used to monitor the temperatures at different locations of the HPCP as labeled in Fig. 4. In order to accurately measure the effective heat transfer rate through the heat pipe, high accuracy (±0.02 °C) platinum resistance thermometers (PRTs) were used to measure the inlet and outlet water temperatures. Before the test, the heat pipe was maintained at 25 °C by the cooling water. After a heat load is applied on the heat pipe evaporator section, the

Fig. 3. Schematic of the cooling loop of the experimental setup.

896

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

Fig. 4. Schematic of the experimental setup for characterizing the HPCP.

system was allowed to reach steady-state when the variation in wall temperature is less than 0.1 °C for 2 min. Experiments were firstly conducted under steady-state conditions to determine the temperature distribution on the HPCP and the most influential factors on the thermal performance of the HPCP. The HPCP was tested with different heating powers (25 W, 50 W, 75 W, 100 W, and 125 W) supplied by the plate heaters. The testing was performed continuously with a 25 W step increment. The condenser sections of the heat pipes of the HPCP were cooled by 25 °C water from a thermostatic water bath with a flow rate of 2 L min1. Fig. 5(a) shows the maximum temperature and temperature difference at the evaporator of the heat pipes. When the heating power was below 75 W, the maximum temperature at the evaporator was controlled below 40 °C, and the temperature difference was controlled below 5 °C. The cooling capacity of the baseline HPCP studied is therefore below 75 W, and needs improvement before it can meet the design target of 108.8 W. Subsequently, to improve the cooling capacity, sensitivity studies were conducted by varying design parameters, including the working fluid of heat pipes, flow rate, coolant temperature, operating orientation, the number of heat pipes in the HPCP and with/ without cooling fins. The results are presented in the following sections. 3.1. Working fluid of heat pipes As heat pipes are the most critical components in a HPTMS, selecting a suitable heat pipe is the primary step in designing such a system. The wick structure of the heat pipes in this study is chosen to be sintered copper powder wick [40] as the heat source orientation and gravity have less effect on such heat pipes as the wicks have stronger capillary action compared to groove and mesh wick [41,42]. Besides, the performance degradation on maximum heat transfer rate of sintered powder heat pipes when flattened is smaller than grooved heat pipes [42]. Tubular heat pipes are widely available in the past decades and have been widely used in many industrial applications for their

Fig. 5. Comparison of the thermal performance of HPCPs with different working fluids (a) evaporator temperature; (b)-(c) condenser temperature.

efficient cooling and thermal management [25,26,43,44]. The selection of heat pipes for a particular thermal management application is based on the cooling power level and the range of evaporator temperatures. For the HPTMS conceptual design described in Section 2, an appropriate heat pipe candidate should have a maximum heat transfer rate of over 27.2 W with a thermal resistance as low as possible. Experimental tests were carried out on various types of heat pipes (outer diameter: 6.0 mm and 8.0 mm; flattened thickness of the evaporator: 2.5 mm, 3.0 mm, 3.5 mm, 4.0 mm; working fluid: water, methanol, ethanol) following a similar testing procedure used by Huang et al. [45]. From the experimental parametric study conducted, it was concluded that an 8.0 mm outer diameter water heat pipe with the evaporator section flattened to 3.0 mm thickness is a good tradeoff between heat pipe performance and size for the current application. With the same configuration, a water heat pipe may have a lower thermal resistance than a methanol heat pipe, however, the higher operating temperature range of a water heat pipe is unfavorable for the application on thermal management of batteries. In view of this conflict, the selection between water heat pipes and methanol heat pipes based on the existing knowledge and datasheets is not straightforward. Experiments were therefore conducted to compare these two candidates at the HPCP level. The results are presented in Fig. 5(a) which showed that the HPCP with water heat pipe has slight advantages over the HPCP with methanol heat pipe in terms of the maximum temperature (Tmax) and temperature difference (Tdiff) across the evaporators. This can be ascribed to the fact that the heating power was too low to fully activate the water heat pipe. This is supported by the transient temperature rises at the heat pipe condenser shown in Fig. 5 (b) and (c). In the cross-flow cooling arrangement, when a heat pipe is fully activated, the temperature distribution at the condenser section should be approximately uniform. For the HPCP utilizing methanol heat pipes shown in Fig. 5(c), there was little difference between T2 and T3, indicating a relatively uniform temperautre distribution along the condenser and a fully activated working condition. In contrast, for the HPCP utilizing water heat

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

pipes as shown in Fig. 5(b), there was a large temperature difference between T2 and T3 at low heat loads and this difference narrowed down with the increase of heat load, showing that the water heat pipes were gradually becoming fully activiated. At the heating power of 100 W and 125 W, the water heat pipe was fully activated, and the maximum temperature of the HPCP utilizing water heat pipes was slightly lower than that of the HPCP utilizing methanol heat pipes. These results suggest that the methanol heat pipe is comparable to the water heat pipe for the present application but the latter is slightly better. 3.2. Coolant flow rate, coolant temperature, operating orientation and the number of heat pipes in a HPCP In this section, parametric studies on the HPCP were conducted and presented in Fig. 6. The experimental parametric studies were performed by varying the coolant flow rate (Fig. 6(a)), coolant temperature (Fig. 6(b)), operating orientation (Fig. 6(c)), and the number of heat pipes in a HPCP (Fig. 6(d)). First, by fixing the number of heat pipes to 4 in the HPCP, the effect of coolant flow rate on the performance of the HPCP was investigated. The results are plotted in Fig. 6(a) which shows that there are large decreases in the maximum temperature and temperature difference at the evaporator when the flow rate is increased from 1 L min1 to 2 L min1. The improvement becomes minimal from 2 L min1 to 3 L min1. This suggests that 2 L min1 may be an optimum flow rate for the present design. Having fixed the flow rate, further investigations were performed on the effect of coolant temperature (15 °C, 20 °C and 25 °C). As can be seen from the results in Fig. 6(b), the decreases in the maximum temperatures at the evaporator are approximately the same as the decreases in the coolant temperature. The maximum temperature was controlled below 40 °C with a coolant temperature of 15 °C when the heating load was 100 W. However, a distinct drawback is the increased temperature difference on the cold plate when lowering the temperature of the coolant. When the heating power was 125 W and coolant inlet temperature was 15 °C, the temperature difference exceeded 5 °C, which is unfavorable for battery operation. Another drawback is

897

that the low initial battery temperature (15 °C for example) may reduce the output power and energy of the battery [4] and shorten the battery cycle life [46–51]. Another critical design parameter is the heat pipe operating orientation, which has been a key concern of many researchers. To evaluate the sensitivity of the performance to the variation of operating orientation, 3 typical orientations were tested: 1. Condenser at the bottom and evaporator on top (+90° orientation); 2. Horizontal (0° orientation); 3. Condenser on top and evaporator at the bottom (90° orientation). The results are presented in Fig. 6(c) which shows that the 90° operating orientation is the best for optimal performance of the HPCP. This result is in agreement with that of Rao et al. [9] The last design parameter investigated in this sensitivity study is the number of heat pipes in a HPCP. In the experiment, HPCPs with different numbers (3, 4, and 5) of heat pipes were assembled and tested. The maximum temperature and temperature difference at the evaporator are plotted in Fig. 6(d) which shows that the maximum temperature decreases with an increasing number of heat pipes in a HPCP, and the temperature difference also decreases slightly with an increase in the number of heat pipes. However, the HPCP with 5 heat pipes which showed the optimum performance was still unable to maintain the maximum temperature below 40 °C when the heating power was 100 W. This result suggests that simply increasing the number of heat pipes in a HPCP to 5 is not sufficient to meet the design requirement. The results in this section have shown that varying the design parameters mentioned above are not sufficiently effective in enhancing the performance of the HPCP. Hence, we now have to consider other solutions to enhance the performance of the HPCP. 3.3. The effect of cooling fins Since the use of cooling fins has been one of the most effective and common techniques in enhancing heat transfer [52–54], experiments were conducted to study the effectiveness of cooling fins in enhancing the performance of a HPCP. As shown in Fig. 4(a), 9 copper fins of dimensions 65 mm
20 mm
0.5 mm were force-fitted onto the condensers of the heat

Fig. 6. Sensitivity studies on (a) flow rate; (b) coolant temperature; (c) operating orientation; (d) the number of heat pipe in a HPCP.

898

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

pipes with a fin pitch of 10 mm. Thereafter, the fins were soldered to the heat pipes to ensure good thermal contact and to minimize the overall thermal resistance of the finned HPCP. The comparisons of the maximum temperature and temperature difference between HPCPs with and without fins are shown in Fig. 7. The maximum temperatures at the evaporator shown in Fig. 7 were significantly reduced when copper fins were added, and specially, the decreases in maximum evaporator temperature was 6.8 °C when the heating power was 100 W with a horizontal orientation, and the maximum evaporator temperature was controlled below 40 °C. As for the temperature differences across the evaporator shown in Fig. 7, adding copper fins also showed a great improvement on the HPCP performance. The temperature difference was brought down to about 1/4 of that without copper fins. Thus it can be confirmed that adding fins on the condensers of heat pipes is an effective approach in reducing the maximum temperature and the temperature difference at the evaporator.

Fig. 7. Comparison of the thermal performance of the HPCP with and without fins.

It should be noted that the performance of the present design is far superior to that of Rao et al. [9]. Despite the lower coolant temperature (21 °C) used by Rao et al. compared to the present test (25 °C), the cooling capacity of the present design at 125 W (Fig. 7), is much higher than the 30 W of Rao’s design, when their maximum temperature is limited to 50 °C and their temperature difference limited to 5 °C.

4. Thermal response of battery cells with the HPTMS 4.1. Experimental tests In the parametric experiments described in the previous section to characterize the performance of the HPCP, plate heaters were used to simulate the heat load from the battery cells. Having optimized the performance of the HPCP, it is still necessary to assess its performance with the integration of battery cells during charging. The same test rig (shown in Fig. 3) which was used to characterize the performance of the HPCP was used for this experiment by replacing the plate heaters with two prismatic cells as shown in Fig. 8(a). One side of each cell (called ‘‘cooled surface”) was attached to the HPCP and the other side (called ‘‘insulated surface”) was insulated by a Styrofoam block to prevent heat loss from the cells to the ambient air. A layer of thermal grease (approximately 1 mm in thickness, thermal conductivity of 3 W m1 K1) was applied at the interface between the HPCP and the cell to ensure good thermal contact. 12 T-type thermocouples were attached to the 2 cells, 3 on each side-surface of each cell, to monitor the transient thermal response of the battery cells. The locations of the 3 thermocouples (T1, T2, T3) on each cell surface are shown in Fig. 8 (a) where T1 and T3 are near the ends of the cells and T2 is at the center of the side-surfaces of the cells. All the tests described in this Section were performed with a coolant flow rate of 2 L min1 and coolant inlet temperature of 25 °C. In the experiments, the cells were charged at various C_rates using a battery charger (Maccor, Inc.) and the temperatures registered by the 12 thermocouples recorded in a data acquisition system.

Fig. 8. Experimental set up for characterizing the performance of the HPTMS, (b) actual battery pack.

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

A typical set of results corresponding to a charging rate of 8 C_rate is shown in Fig. 9. As can be seen from Fig. 9, the temperatures of T1 and T3 on the insulated surface of Battery Cell 1 were found to be slightly lower than that of T2 at the central point. This is due to greater heat loss through the ends of the battery cells. Similar differences between T1, T3, and T2 were observed for the other 3 surfaces as well, which have not been plotted in Fig. 9 for reason of clarity. It was also observed that the temperature distributions along the 2 insulated surfaces of the two battery cells were very similar, and those along the 2 cooled surfaces as well. This can be seen in the closeness of the values of T2 for the 2 insulated surfaces and the 2 cooled surfaces plotted in Fig. 9. Since T2 is less affected by heat loss through the ends of the battery cells, it is a more accurate indicator of the cell surface temperature. Hence in the following discussions T2 will be used as a measure of the battery cell temperature. The experimental results for different C_rates of charging are presented in Fig. 10. As can be seen, the temperatures of the cooled

899

surfaces were maintained below 38 °C during 3 C_rates, 5 C_rates and 8 C_rate charging tests. Although the temperatures of the insulated surfaces exceeded 40 °C during 5 C_rate and 8 C_rate charging, it should be noted that in the actual battery pack, both side-surface of the cell will be cooled by a HPCP (Fig. 8(b)) so that their temperature will not be as high as that measured in the experiment. In fact, the heat load handled by the HPCP in these experiments is about double that in the actual battery pack. This is because in the experiment here, the 2 cells and the HPCP are completely wrapped with insulation to minimize heat loss (Fig. 8 (a)). Hence the HPCP in the experiment is subjected to the full heat load from 2 cells, whereas in the actual battery pack (Fig. 8(b)) both the side surfaces of each cell will be in contact with a HPCP so that the heat load for each HPCP is equivalent to the heat load from one cell only. The fact that the temperature of the cooled battery cell surface in the experiment is kept below 38 °C despite the doubling of the heat load demonstrates that the HPCP designed can in fact handle twice the rated heat load from the battery cells. This is consistent with the safety factor of 2 which have been used in the design of the HPTMS. 4.2. Numerical simulations

Fig. 9. Temperature rise at various locations on the surfaces of the cells.

Fig. 10. Experimental results on Transient rise of cell temperature during 3 C_rate, 5 C_rate and 8 C_rate of charging.

In order to estimate the magnitude of temperature rise in the actual battery pack, a 3-dimensional thermal model was developed and solved numerically to predict the transient temperature rise of cells in the actual battery pack. For the purpose of comparison and also to verify the numerical model with experiment, a similar thermal model was developed and solved numerically for the experimental set up presented in Section 4.1. Taking symmetry into account, the problem domain for the numerical simulation of a battery cell in the experimental set up is a prismatic cell with a convective boundary condition on the cooled surface which is in contact with the HPCP and an adiabatic boundary condition on the insulated surface (Fig. 11(a)) which is thermally insulated by a Styrofoam insulator. For the numerical simulation of a cell in the battery pack, the problem domain is half a cell shown in Fig. 11(b) with a convective boundary condition on the cooled surface which is in contact with the HPCP and a symmetry boundary condition on the central plane of the battery cell. For the convective boundary conditions at the surfaces in contact with the HPCP, the effective heat transfer coefficient h with respect to a HPCP coolant temperature of Tf = 25 °C was estimated from Eq. (1) below using the experimental data from steady-state tests conducted on the HPCP.

Fig. 11. Boundary conditions for the numerical simulations (a) cell in experimental setup (b) cell in a pack.

900

h¼

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

Q eff AðT w  T f Þ

ð1Þ

where Q eff is the effective heat transfer rate which can be calculated from the sensible heat gain of the cooling water in the HPCP characterization experiment (Section 3.3), A is the contact area between the HPCP and the plate heater, and Tw is the measured average temperature of the HPCP. The governing differential equation for the simulation is the energy equation,

qC p

@T ¼ r  ðkrTÞ þ Q @t

ð2Þ

where q is the density of the battery cell (2110 kg m3), Cp is the specific heat capacity (1264 J kg1 K1), k is the thermal conductivity (kx = 1.4 W m1 K1, ky = kz = 15.6 W m1 K1), and Q is the transient heat generation rate which had been measured using an adiabatic calorimeter (THT., Ltd) when the battery cell was charged by a battery charger (Maccor, Inc,) at the appropriate C_rate. The measured results presented in Fig. 2 [39] were interpolated as functions of time and imported as the transient, uniform heat sources in the battery cells in the simulations. To account for the effect of thermal grease on the cell temperature, an effective thermal resistance (3.75
104 m2 K W1) was assigned to each cooled surface. The numerical simulations were performed using the heat transfer module of a commercial finite element solver, COMSOL Multiphysics 4.3b. The model was computed using the GMRES solver. A tight convergence tolerance of 1.0
106 was set for the governing equations in all simulations. In addition, the grid independent test was carried out to refine the grid size until the results are not affected by further refinement of the mesh and the error of the results is kept within 5%. In this modeling effort, thermal radiation transfer was assumed to be negligible and was not taken into account. The simulated temperatures at the battery cell surfaces for the experimental set up are compared with the experimental values in Fig. 12(a). As can be seen, the simulated results matched well with the experimental measurements with a mean absolute

Fig. 12. Transient rise of cell temperature during 3 C_rate, 5 C_rate and 8 C_rate of charging (a) measured and simulated for experimental set up; (b) simulated for actual battery pack.

error of less than 3.6% for each temperature curve. Such a good agreement verifies the accuracy of the numerical simulations, and lays the basis for the following prediction of the thermal response of the cells in the actual battery pack. As shown in Fig. 12(b), when a cell in the pack is subjected to cooling on both sides by HPCPs, the internal temperature of the battery cell could be controlled to below 45 °C at the end of an 8 C_rate current charging process. This suggests that the present HPTMS roughly meets the requirements of battery thermal management. To further enhance the performance of the present HPTMS so as to meet the tight requirement on maintaining the battery cell temperature below 40 °C, several methods like increasing the flow rate or increasing the number of cooling fins could be implemented. In the below section, the influence of different cooling strategies on the thermal response of the battery cell will be investigated.

5. Experimental study on different cooling strategies In this section, thermal responses of battery cells subjected to 5 different cooling strategies were investigated during 8 C_rate charging. The cooling strategies are: 1. Constant 25 °C: 25 °C of cooling water constantly flows through the system. 2. Constant 15 °C: 15 °C of cooling water constantly flows through the system. 3. Decreasing: cooling water is supplied with a flow rate of 2 L min1, the inlet temperature of the coolant gradually decreases from 25 °C when the charging is initiated. 4. Quench: The temperature of the system with battery cells is initially maintained at 25 °C and the gate valve (Fig. 3) of the cooling loop is closed, the temperature of water bath is maintained at 15 °C and the gate valve is opened when the charging is initiated. 5. Delay quench: The temperature of the system with battery cells is initially maintained at 25 °C and the gate valve of the cooling loop is closed, the temperature of water bath is maintained at 15 °C and the gate valve is opened 100 s after the charging is initiated. The flow rate of the above cooling strategies is 2 L min1, and temperature responses at the inlet of the cooling channel in different cooling strategies were presented in Fig. 13(a). In the case of ‘‘constant 25 °C” and ‘‘constant 15 °C”, the inlet coolant temperature was kept at the 25 °C and 15 °C, respectively. In the case of ‘‘decreasing”, the inlet temperature kept decreasing, however, the decreasing rate is limited by the cooling capacity of the water bath, and the coolant temperature at the end of charging process was about 21 °C. In the case of ‘‘quench”, the temperature went through a sharp decrease to 17 °C due to the mixing effect of the coming coolant (15 °C) and the existing water (25 °C) in the cooling channel. The coolant temperature kept decreasing because the refrigerant loop of the water bath kept functioning. The coolant temperature in the case of ‘‘delay quench” underwent a similar pattern as the case of ‘‘quench” except for a 100 s delay in the temperature drop. The transient temperature responses at the cooling surface and the insulated surface were presented in Fig. 13(b) and (c), respectively. As aforementioned, it has been stressed by many researchers that lithium-ion batteries are more prone to lithium platting at low temperature due to the sluggish electrochemistry [46–50], and this issue becomes more serious when it comes to fast charging using high C_rate current. The monitored battery cell charging voltages as shown in Fig. 13(d) showed that the charging potential of battery cells with ‘‘constant 15 °C” cooling has a higher voltage

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

901

Fig. 13. (a) Monitored coolant temperature at the inlet; (b) temperature evolutions at the cooling surface of the cell; (c) temperature evolutions at the insulated surface of the cell; (d) cell potential during charging.

than that of cells with ‘‘constant 25 °C” cooling, resulting in a higher over-potential and lower charged capacity. The phenomenon is due to the lower lithium ion diffusivity and reaction rate at lower temperature, which has been elucidated by many researchers [4,48–50]. Therefore, to minimize the occurrence of lithium platting due to the low lithium ion intercalation rate at low temperature, a TMS is also recommended to maintain the battery cell temperature above 25 °C. Having the above consideration, other alternative cooling strategies were proposed and investigated. The first strategy investigated is the ‘‘decreasing” cooling, in which the coolant temperature keeps decreasing after the charging is initiated. The temperatures of battery cells in this case showed minor decrements as compared to those in the case of ‘‘constant 25 °C” due to the limitation on slow cooling rate of the water bath and the high heat capacity of the water in the water bath. A potential solution to this issue is to precool the water in the water bath down to 15 °C and leave the cooling loop idle at 25 °C. When the gate valve is opened in the cooling loop, the 15 °C cooling water flashing into the cooling loop quenches the battery cell rapidly. The temperature at the cooling surface was reduced by 7 °C lower than of ‘‘constant 25 °C”. Nevertheless, the deficient is that the coolant was quenched down below 25 °C at the initial charging period. To avoid this overcooled situation, the ‘quench” strategy is improved to ‘‘delay quench” method, which allows a short time lag for the battery cell to be heated up to a higher value around 30 °C. As observed in Fig. 13(b), the temperature of the battery cell at the cooled surface was maintained above 25 °C but was much lower than that of ‘‘constant 25 °C” cooling. To evaluate the thermal response of battery cells subjected to the cooling of the HPTMS at the battery pack level (Fig. 8(b)), similar numerical efforts as the previous section were undertaken. The predicted results as presented in Fig. 14 confirmed that the ‘‘delay quench” cooling and ‘‘decreasing” cooling strategies successfully maintained the cell temperature within the range of 25 °C–40 °C, and the ‘‘delay quench” cooling strategy showed advantage over

Fig. 14. Predicted battery temperature evolutions with different cooling strategies during 8 C_rate of charging.

‘‘decreasing” cooling in maintaining the cell temperature within a lower range. Note that the time lag in the ‘‘delay quench” cooling strategy is an alterable design parameter that can be adjusted from case to case depending on the charging C_rate and the coolant temperature, the detail adjustment would be an interesting topic for future research. 6. Summary In order to deal with the large amount of heat generation within battery cells during fast charging and severe scenarios of EVs, efforts in this work were undertaken to enhance the cooling

902

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903

performance of a HPTMS for prismatic/pouch type lithium-ion batteries. To meet the target, sensitivity studies were performed on different design parameters (flow rate, coolant temperature, operating orientation, the number of heat pipes in a HPCP, working fluid of heat pipes, and the application of cooling fins). The final improved design of the HPTMS was evaluated experimentally with actual battery cells during the charging process. The results confirmed the feasibility and effectiveness of the improved HPTMS for thermal management of batteries during fast charging. The following are the main findings from the experimental investigation: 1. Water and methanol are comparatively good as heat pipe working fluids for battery cooling. Ethanol is not suitable due to premature dry-out. 2. The performance of the HPTMS is sensitive to the variation of coolant temperature, and the decrease of maximum heat pipe evaporator temperature is approximately the same as that of the decrease in the coolant temperature. However, a low coolant temperature is not recommended for lithium-ion batteries because of lithium plating during fast charging at low temperature. 3. Adding cooling fins at the heat pipe condenser sections to increase the area of heat transfer does significantly increase the cooling capacity of the HPTMS, and is confirmed to be the most effective way in reducing the maximum temperature and temperature difference in the battery. 4. The effectiveness of the HPTMS was evaluated experimentally with actual prismatic cells at different C_rates of charging. It was demonstrated that the HPTMS was well able to handle twice the maximum heat generation from 8 C_rate charging of the batteries. 5. The cooling strategy has great impact on the thermal response batteries. Precooling batteries to a low temperature in the ‘‘constant 15 °C” strategy is detrimental to the batteries because of the lithium plating issue during fast charging at low temperature. The ‘‘decreasing” cooling strategy which gradually decreases the coolant temperature is limited by the cooling rate of the water bath, and the improvement in thermal performance of the HPTMS is insignificant. The ‘‘delay quench” cooling strategy, which quenches the 25 °C of batteries using 15 °C of coolant with a short time lag after the charging process initiates, shows the optimum cooling performance in maintaining the battery temperature within the range of 25 °C–40 °C. Conflict of Interest None declared. Acknowledgements The authors gratefully acknowledge the Grant EPD090005RFP(A) from Energy Market Authority Singapore.

No.

References [1] L.W. Jin, P.S. Lee, X.X. Kong, Y. Fan, S.K. Chou, Ultra-thin minichannel LCP for EV battery thermal management, Appl. Energy 113 (2014) 1786–1794. [2] S.K. Mohammadian, Y. Zhang, Thermal management optimization of an aircooled Li-ion battery module using pin-fin heat sinks for hybrid electric vehicles, J. Power Sources 273 (2015) 431–439. [3] Y. Ye, Y. Shi, A.A.O. Tay, Electro-thermal cycle life model for lithium iron phosphate battery, J. Power Sources 217 (2012) 509–518. [4] Y. Ye, Y. Shi, N. Cai, J. Lee, X. He, Electro-thermal modeling and experimental validation for lithium ion battery, J. Power Sources 199 (2012) 227–238. [5] R. Kizilel, R. Sabbah, J.R. Selman, S. Al-Hallaj, An alternative cooling system to enhance the safety of Li-ion battery packs, J. Power Sources 194 (2009) 1105– 1112. [6] A.A. Pesaran, Battery thermal models for hybrid vehicle simulations, J. Power Sources 110 (2002) 377–382.

[7] T.M. Bandhauer, S. Garimella, T.F. Fuller, A critical review of thermal issues in lithium-ion batteries, J. Electrochem. Soc. 158 (2011) R1–R25. [8] T. Tran, S. Harmand, B. Sahut, Experimental investigation on heat pipe cooling for hybrid electric vehicle and electric vehicle lithium-ion battery, J. Power Sources 265 (2014) 262–272. [9] Z. Rao, S. Wang, M. Wu, Z. Lin, F. Li, Experimental investigation on thermal management of electric vehicle battery with heat pipe, Energy Convers. Manage. 65 (2013) 92–97. [10] Z. Rao, S. Wang, A review of power battery thermal energy management, Renewable Sustainable Energy Rev. 15 (2011) 4554–4571. [11] A. Jarrett, I.Y. Kim, Design optimization of electric vehicle battery cooling plates for thermal performance, J. Power Sources 196 (2011) 10359–10368. [12] H. Park, A design of air flow configuration for cooling lithium ion battery in hybrid electric vehicles, J. Power Sources 239 (2013) 30–36. [13] R. Liu, J. Chen, J. Xun, K. Jiao, Q. Du, Numerical investigation of thermal behaviors in lithium-ion battery stack discharge, Appl. Energy 132 (2014) 288–297. [14] J. Xun, R. Liu, K. Jiao, Numerical and analytical modeling of lithium ion battery thermal behaviors with different cooling designs, J. Power Sources 233 (2013) 47–61. [15] F. He, X. Li, L. Ma, Combined experimental and numerical study of thermal management of battery module consisting of multiple Li-ion cells, Int. J. Heat Mass Transfer 72 (2014) 622–629. [16] F. He, L. Ma, Thermal management of batteries employing active temperature control and reciprocating cooling flow, Int. J. Heat Mass Transfer 83 (2015) 164–172. [17] A.A. Pesaran, Battery Thermal Management in EVs and HEVs: Issues and Solutions, Advanced automotive battery conference, Las Vegas, Nevada, 2001. [18] A. Pesaran, S. Burch, M. Keyser, An Approach for Designing Thermal Management Systems for Electric and Hybrid Vehicle Battery Packs, Fourth Vehicle Thermal Management Systems Conference and Exhibition, London, UK, 1999. [19] S. Al Hallaj, J.R. Selman, A novel thermal management system for electric vehicle batteries using phase-change material, J. Electrochem. Soc. 147 (2000) 3231–3236. [20] Z. Ling, J. Chen, X. Fang, Z. Zhang, T. Xu, X. Gao, S. Wang, Experimental and numerical investigation of the application of phase change materials in a simulative power batteries thermal management system, Appl. Energy 121 (2014) 104–113. [21] N. Javani, I. Dincer, G.F. Naterer, G.L. Rohrauer, Modeling of passive thermal management for electric vehicle battery packs with PCM between cells, Appl. Therm. Eng. 73 (2014) 307–316. [22] N. Javani, I. Dincer, G.F. Naterer, B.S. Yilbas, Heat transfer and thermal management with n{PCMsn} in a Li-ion battery cell for electric vehicles, Int. J. Heat Mass Transfer 72 (2014) 690–703. [23] X. Yang, Y.Y. Yan, D. Mullen, Recent developments of lightweight, high performance heat pipes, Appl. Therm. Eng. 33 (2012) 1–14. [24] K. Nithyanandam, R. Pitchumani, Design of a latent thermal energy storage system with embedded heat pipes, Appl. Energy 126 (2014) 266–280. [25] C.A. Chung, S. Yang, C. Yang, C. Hsu, P. Chiu, Experimental study on the hydrogen charge and discharge rates of metal hydride tanks using heat pipes to enhance heat transfer, Appl. Energy 103 (2013) 581–587. [26] Y. Weng, H. Cho, C. Chang, S. Chen, Heat pipe with PCM for electronic cooling, Appl. Energy 88 (2011) 1825–1833. [27] J. Du, P. Bansal, B. Huang, Simulation model of a greenhouse with a heat-pipe heating system, Appl. Energy 93 (2012) 268–276. [28] E. Firat, G. Bandlamudi, P. Beckhaus, A. Heinzel, Heat Pipe Assisted Thermal Management of a HT PEMFC Stack, European COMSOL Conference, Milan, Italy, 2012. [29] A. Faghri, Heat Pipe Science and Technology, Taylor & Francis, 1995. [30] V. Maziuk, A. Kulakov, M. Rabetsky, L. Vasiliev, M. Vukovic, Miniature heatpipe thermal performance prediction tool-software development, Appl. Therm. Eng. 21 (2001) 559–571. [31] M.S. Wu, K.H. Liu, Y.Y. Wang, C.C. Wan, Heat dissipation design for lithium-ion batteries, J. Power Sources 109 (2002) 160–166. [32] A. Greco, D. Cao, X. Jiang, H. Yang, A theoretical and computational study of lithium-ion battery thermal management for electric vehicles using heat pipes, J. Power Sources 257 (2014) 344–355. [33] Q. Wang, B. Jiang, Q.F. Xue, H.L. Sun, B. Li, H.M. Zou, Y.Y. Yan, Experimental investigation on EV battery cooling and heating by heat pipe, Appl. Therm. Eng. 1–7 (2014). [34] R. Zhao, J. Gu, J. Liu, An experimental study of heat pipe thermal management system with wet cooling method for lithium ion batteries, J. Power Sources 273 (2015) 1089–1097. [35] L.H. Saw, Y. Ye, A.A.O. Tay, Electrochemical-thermal analysis of 18,650 Lithium Iron Phosphate cell, Energy Convers. Manage. 75 (2013) 162–174. [36] C. Botsford, A. Szczepanek, Fast Charging vs. Slow Charging: Pros and Cons for the New Age of Electric Vehicles, EVS24, Stavanger, Norway, 2009. [37] Y. Ye, L.H. Saw, Y. Shi, K. Somasundaram, A.A.O. Tay, Effect of thermal contact resistances on fast charging of large format lithium ion batteries, Electrochim. Acta 134 (2014) 327–337. [38] P. Nemec, A. Caja, M. Malcho, Mathematical model for heat transfer limitations of heat pipe, Math. Comput. Model. 57 (2013) 126–136. [39] Y. Ye, Y. Shi, L.H. Saw, A.A.O. Tay, Numerical analyses on a heat pipe thermal management system for lithium-ion batteries during fast charging, Appl. Therm. Eng. 86 (2015) 281–291.

Y. Ye et al. / International Journal of Heat and Mass Transfer 92 (2016) 893–903 [40] R. Manimaran, K. Palaniradja, N. Alagumurthi, J. Hussain, Factors affecting the thermal performance of heat pipe–a review, J. Energy Res. Stud. 3 (2011) 20–24. [41] C.K. Loh, E. Harris, D.J. Chou, Comparative study of heat pipes performances in different orientations, in: Semiconductor Thermal Measurement and Management Symposium, 2005 IEEE Twenty First Annual IEEE, San Jose, CA USA, 2005, pp. 191–195. [42] K. Lin, S. Wong, Performance degradation of flattened heat pipes, Appl. Therm. Eng. 50 (2013) 46–54. [43] K.H. Do, S.J. Kim, S.V. Garimella, A mathematical model for analyzing the thermal characteristics of a flat micro heat pipe with a grooved wick, Int. J. Heat Mass Transfer 51 (2008) 4637–4650. [44] S.J. Kim, J.K. Seo, K.H. Do, Analytical and experimental investigation on the operational characteristics and the thermal optimization of a miniature heat pipe with a grooved wick structure, Int. J. Heat Mass Transfer 46 (2003) 2051– 2063. [45] L. Huang, M.S. El-Genk, J.M. Tournier, Transient performance of an inclined water heat pipe with a screen wick, in: A. Faghri, A.J. Juhasz, T. Mahefkey (Eds.), 1993 ASME National Heat Transfer Conference, Atlanta, GA, Heat Pipes Capillary Pumped Loops HTD 236 (1993) 87–92. [46] A.N. Jansen, D.W. Dees, D.P. Abraham, K. Amine, G.L. Henriksen, Lowtemperature study of lithium-ion cells using a LiySn micro-reference electrode, J. Power Sources 174 (2007) 373–379.

903

[47] M. Ouyang, Z. Chu, L. Lu, J. Li, X. Han, X. Feng, G. Liu, Low temperature aging mechanism identification and lithium deposition in a large format lithium iron phosphate battery for different charge profiles, J. Power Sources 286 (2015) 309–320. [48] S. Tippmann, D. Walper, L. Balboa, B. Spier, W.G. Bessler, Low-temperature charging of lithium-ion cells part I: Electrochemical modeling and experimental investigation of degradation behavior, J. Power Sources 252 (2014) 305–316. [49] M. Petzl, M. Kasper, M.A. Danzer, Lithium plating in a commercial lithium-ion battery – a low-temperature aging study, J. Power Sources 275 (2015) 799– 807. [50] J. Fan, S. Tan, Studies on charging lithium-ion cells at low temperatures, J. Electrochem. Soc. 153 (2006) A1081–A1092. [51] J. Wang, P. Liu, J. Hicks-Garner, E. Sherman, S. Soukiazian, M. Verbrugge, H. Tataria, J. Musser, P. Finamore, Cycle-life model for graphite-LiFePO4 1cells, J. Power Sources 196 (2011) 3942–3948. [52] V.W. Khond, Others, heat transfer enhancement in fin and tube heat exchanger-a review, J. Eng. Appl. Sci. 8 (2013). [53] F.P. Incropera, Fundamentals of Heat and Mass Transfer, John Wiley & Sons, 2011. [54] R.L. Webb, N. Kim, Principle of enhanced heat transfer, 1994.