An innovative practical battery thermal management system based on phase change materials: Numerical and experimental investigations

Accepted Manuscript Research Paper An innovative practical battery thermal management system based on phase change materials: Numerical and experimental investigations Amine Lazrak, Jean-François Fourmigué, Jean-François Robin PII: DOI: Reference:

S1359-4311(17)31573-9 http://dx.doi.org/10.1016/j.applthermaleng.2017.08.172 ATE 11055

To appear in:

Applied Thermal Engineering

Received Date: Revised Date: Accepted Date:

7 March 2017 21 July 2017 31 August 2017

Please cite this article as: A. Lazrak, J-F. Fourmigué, J-F. Robin, An innovative practical battery thermal management system based on phase change materials: Numerical and experimental investigations, Applied Thermal Engineering (2017), doi: http://dx.doi.org/10.1016/j.applthermaleng.2017.08.172

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

An innovative practical battery thermal management system based on phase change materials: Numerical and experimental investigations Amine Lazraka, Jean-François Fourmiguéb, Jean-François Robinb a

Concordia University, Department of building civil and environmental engineering, Montreal, Canada b

CEA, LITEN, INES, Le Bourget du Lac, France

Corresponding Author: [email protected]

ABSTRACT

The market of electric vehicles still faces some impediment to its optimal development. Electric batteries play an important role in this context since they are the key element in an electric vehicle (EV). Improving the energy performance of batteries will certainly improve the autonomy and reliability of EVs and thus their market penetration. To achieve this objective, battery thermal management systems are necessary to keep the temperature below security limits and make the temperature distribution as uniform as possible inside the battery pack and its cells. In this paper, a new solution to integrate and improve the thermal heat transfer of a phase change material (PCM) inside a battery thermal management system (BTMS) is proposed and the effect of the PCM melting temperature on the system performance is investigated. Two numerical models have been built and their results were the input of a small size PCM-based BTMS prototype development. Experimental results showed that the novel system was able to reduce the system temperature by at least 5°C, compared to the reference, upon completion of the phase change process inside the PCM.

KEYWORDS

Thermal management, Electric battery, Phase change material, Modelling, System development, System testing.

Nomenclature Heat power Volume Density Temperature Thermal capacity Total enthalpy Velocity field Heat flux Thermal conductivity Latent heat Liquid fraction Sensible enthalpy Convection heat transfer coefficient Front surface position Time PCM thermal diffusivity Subscripts

[W] [m3] [kg.m-3] [°C] [J.kg-1.K-1] [J. m-3] [m.s-1] [W.m-2] [W.m-1.K-1] [J.kg-1] [-] [J. m-3] [W.m-2.K-1] [m] [s] [m2.s-1]

Melting Heat source (cell) Cell (or fake cell) PCM Liquid Solid

Abbreviations PCM BTMS EV 1D 3D

TC 1.

Phase change material Battery thermal management system Electric vehicle One space dimension modelling Three space dimensions modelling Reynolds number Nusselt number Thermo-couple

Introduction

Nowadays, the environment protection received considerable attention in global scale and a priority for the majority of nations. In this context, considering the fact that vehicles are responsible for a large amount of CO2 emissions (almost 10%) [1], electric mobility can play an important role in the environment preservation. As the fossil energy resources are limited and their price keep growing, hybrid electric vehicles (EVs) are an alternative solution to conventional thermal ones. Global deployment of EV battery charging, using electricity generated by centralized and non-centralized renewable systems, is also another way of taking advantage from the electric mobility to reduce fossil energy consumption and greenhouse gas emission. Advantages of EVs include their silent operation, fast reactivity and ease to control. Moreover, EVs can be used as an intermediary storage for demand side electricity management to contribute to performance enhancement of smart grids [2], [3]. For all of these reasons, several countries are putting efforts in research programs and provide aids through subventions in order to make hybrid and EVs competitive and reliable. However, the EV market still faces some serious impediment to its development due to the low autonomy and reliability of the electric batteries [4].

Performance improvement and reliability of EVs highly depends on the electric batteries performance and reliability. Different electrochemical rechargeable batteries are used in EVs but Lithium-ion batteries are the most common ones due, notably, to their high density and long life cycle. A high power is usually needed during the EV operation, at least 10kW for small-size vehicles. To generate such a large power, several cell accumulators are combined in parallel or in series in what is usually called a battery pack. This combination results in generating more and more thermal energy due to Joule effect or exothermic chemical reactions. Being aware of its importance, the research community has focused a lot on this phenomenon. Ramadass et al. in Ref. [5] have shown that after 800 cycles of discharging of a battery at 50°C, it losses 60% of its initial capacity and 70% after 600 cycles at 55°C. During the acceleration phase, the energy extracted from the battery always increases, especially while climbing a slope. As a result, the temperature of the battery increases sharply. Battery temperature fluctuations (due to the environment or acceleration) and its non-uniformity lead to very high local temperature points that are usually responsible for its performance deterioration and thermal runaway [6]. It is not infrequent that some companies recall a high number of electric mobility systems due to this problem [7]. According to the previous paragraph, controlling the battery temperature appears to be mandatory in order to prevent thermal runaway, prolong the battery life cycle and enhance the energy performance and security. Some of the main tasks [8] of a battery thermal management system (BTMS) are: i)

to keep the battery temperature within the constructor requirement bounds;

ii)

to maintain the temperature uniform. Also, the BTMS should be as compact as possible. There are two basic types of BTMS, e.g. passive and active systems. The former uses the environment of the battery to control its temperature (natural ventilation, specific materials, etc.), while the latter utilizes an active system to control the temperature (ventilation, hydraulic loop). The combination of active and passive systems is also of interest.

Among the available methods for thermal management in batteries, PCM-based BTMS shows the most promising performance. In fact, PCM-based BTMS, makes the best trade-off between cost, integration facility, efficiency, life cycle and maintenance costs [8]. Classical active methods, such as air ventilation, are usually not able to achieve uniform temperature along the batteries and accumulators [6], especially for fast discharge phases. This affects the battery pack performance negatively. Moreover, other methods are not reliable and need continuous maintenance. It is important to note that the heat generated in a battery at high discharge currents usually fluctuates. Therefore, it is difficult for the ventilation system to follow this instantaneous evolution. As a result, utilization of a control system, electronic circuits, etc. seems to be inevitable, adding to the complexity of the system. PCM integration to a battery is an innovative way for thermal management. In the last six years, PCM integration solutions have been suggested by the research community [9] [10] [11] [12] [13]. However, there are still some challenges that need to be addressed prior to the commercialization of such systems for EV applications. As its name shows, a PCM is a material that undergoes phase change within a small temperature range (10-80°C) from solid to liquid or vice versa. During this transformation, the PCM absorbs or releases a large amount of energy compared to conduction heat exchange. The advantage of using PCM in a BTMS is its ability to store thermal energy, produced in the battery, at an almost constant temperature (its melting temperature) as long as it does not completely change its phase. Thus, a PCM helps keeping the temperature uniform and constant at the desired level. To achieve this goal, the PCM is usually put around each cell.

Different PCMs with different characteristics are available in the market. For each application, depending on the maximal temperature allowed by a battery, it is possible to find a PCM whose molten temperature is within the required range. This makes PCMs perfectly suitable for different kind of batteries. A list of commonly used PCMs is presented in [8]. However, PCMs have low thermal conductivity values (usually less than 2 W.m-1.K-1). As a consequence, when the PCM completely changes its phase, the battery temperature highly increases. Therefore, it is important to find methods to increase PCM conductivity and delay the complete phase change. Conventional thermal management systems (ventilation or natural cold air, hydraulic loops, etc.) are usually voluminous, complicated and very expensive [14]. That is why there is a need for a new thermal management solution. Using BTMS based on PCM was initially proposed by Al-Hallaj and Selman in Ref. [7] and patented by the same authors in 2000. The researchers have shown the great interest of using PCM to thermalize EV batteries. Thereafter, in the last six years, some studies have tackled the same subject. In the following, a brief description of the most relevant experimental investigations is presented. Khateeb et al. in Ref. [15] compared different heat dissipation systems of a scooter battery: (i) natural convection cooling; (ii) integration of a matrix with aluminum foam to enhance the heat transfer; (iii) PCM based BTMS and (iv) a combination of the PCM and the aluminum foam. Their experiments showed that the method (iv) had the best results. Following the comparison of various experimental results, the authors announced the need to focus on increasing the thermal conductivity of pure PCMs. The use of PCMs provides high efficiency in terms of performance and safety in very demanding conditions of electric current and high temperature. In fact, a BTMS based on PCM allows the battery to operate up to 90% of its capacity under extreme conditions [16], and it also helps stopping the spread of thermal runaway to adjacent accumulators in case of a failure. The results of the study conducted by Sabbah et al. in Ref. [6] showed that a BTMS based on PCM was more efficient than forced air flow cooling. In fact, thanks to its ability to store heat during the phase change process, the PCM significantly prevented the temperature raise inside the accumulator, in normal or abnormal operation. In addition to that, the PCM made the temperature distribution uniform over the battery. Other studies have shown the importance of enhancing the conductivity of PCMs [17] [18] [19] [20]. This helps speeding up the heat transfer from the battery to the PCM. In this context, Li et al. in Ref. [21] used a porous copper foam saturated with paraffin. The results showed that a PCM-based BTMS was clearly advantageous compared to natural convection. The comparison between a pure PCM-based BTMS and a composite PCM-based BTMS showed that the two solutions were able to reduce the system temperature drastically. For the considered period of time, the composite PCM-based BTMS further reduced the battery temperature by about 5°C. However, without an auxiliary ventilation system, it was difficult to judge which BTMS system had higher efficiency. In fact, using the metal foam reduced the available amount of PCM and thus the thermal storage capacity. For long periods of operation, the battery temperature with the composite PCM-based BTMS would raise faster in the case of pure PCM-based BTMS. For this reason, a specific attention should be given to design the BTMS towards the operation periods. Other studies have shown also the interest of enhancing the system thermal conductivity by adding some graphite sheets among cells and composite PCM in Ref. [22] [23] or aluminum foam in Ref. [20]. Z. Ling et al. in Ref. [24] investigated a hybrid BTMS combining PCMs with forced-air cooling. The designed system was able to effectively maintain the battery temperature in the acceptable bounds. In one hand, such hybrid systems are the most relevant ones but, in the other hand, they are more complex and have limitations, when the discharge rates are high, due to the PCMs low thermal conductivity that prevents battery cooling. The analysis presented in previous paragraphs shows the high interest of using a PCM for thermal management in batteries. Different methods have been proposed to overcome the PCM integration limitations in BTMS. Based on the literature review, two major aspects of its integration still need some further investigation: effective thermal conductivity enhancement while

maximizing the PCM mass and PCM regeneration after complete melting. Despite the efficiency of the proposed solutions in the literature, their implementation is still challenging and expensive. In fact, methodologies to enhance the PCM thermal conductivity such as using a composite PCM and copper foam are not simple and need specific implementation and production tools to select the optimal porosity size, shape, composite material, etc. Furthermore, for the case of composite materials, further investigations should be carried out in order to make sure all security requirements are filled for a BTMS application. The objective of this paper is to present the development process, from modelling to small-size prototype tests, of an innovative PCM-based BTMS which is described step by step. The proposed process first highlights the importance of PCM selection (thermal conductivity, melting temperature) and presents the numerical models to be used for optimal sizing. This paper tackles as well an important point which is the effect of the PCM melting temperature, on the system performance, which was not thoroughly investigated previously [25] [26]. Thereafter, the PCM thermal conductivity enhancement combined with a simple ventilation system are carried out. The proposed method is a simple but efficient method to enhance the heat transfer inside the PCM compared to state-of-the-art methods such as metallic foam whose design could be complex and expensive (porosity, weight, etc.). Besides, the proposed solution can reduce the PCM amount to be integrated. 2.

Towards a new practical approach to thermal management

2.1. Methodology A BTMS with high performance should respect the conditions and specifications presented in section 1. To develop such an innovative system using PCM, which is able to overcome the current BTMS limitations, the following development process steps are taken: 1.

1D modelling: Analysis of the effect of PCM thermophysical properties on the BTMS performance to prevent battery overheating more effectively. This step gives recommendations to better select the suitable PCM.

2.

3D modelling: Determination of an improved BTMS design as well as PCM integration in a battery.

3.

Prototype development: Investigation of the performance of the proposed new solution (BTMS design as well as PCM thermal conductivity enhancement method) using a small scale PCM-based BTMS

As mentioned earlier, the proposed process first highlights the importance of PCM selection, the goal includes defining selection criteria of the suitable PCM for BTMS applications. Then, a design concept is presented based on how to enhance the thermal conductivity inside the PCM. This concept is based on using copper dutch weave combined to conventional ventilation to improve the system performance regarding the aforementioned limitations. In this study, several parametric studies using the developed 1D model were carried out to define the optimal physical characteristics of the proposed solution. This simplified 1D model is suitable for parametric studies compared to the 3D one due to the required computational power and time. These studies focused on the thermal conductivity, the melting temperature and the PCM regeneration using forced ventilation. Thereafter, the solution was simulated using a 3D model in order to determine the best integration architecture of the PCM inside a battery pack with different cells. Finally, the successful configuration, which is most practical and efficient, was used to build a small scale BTMS using the improved PCM. 2.2. System modelling 2.2.1. 1D modelling: PCM selection

The developed model enables simulation of heat transfer between the PCM and the cell, which represents the heat source having a cylindrical geometry. This model will be used later to perform a parametric study in order to optimize the PCM characteristics and integration in order to ultimately obtain a thermal management system that meets the specifications. Different approaches have been suggested in the literature to solve phase change problems (latent heat source and apparent heat capacity approaches [27][16]). In this study, the method developed by Voller in Ref. [28][17] was selected as the most suitable to achieve the objectives. In fact, this method, which is based on an enthalpy formulation of the energy balance, has several advantages such as its implementation facility, reduced calculation time, etc.

Figure 1: Scheme of the 1D geometry of the considered problem. The dominant heat transfer mechanism between the cell (where heat is generated:

) and the PCM is by conduction and between the PCM and its environment is by convection.

In this paper, only one cell is considered for simulation as the cells temperature evolution is merely similar considering an insulated battery. In Figure 1, the cell is modelled by a heat source (

) produced by Joule effect. Part of this heat (i.e. sensible

heat) will be stored in the accumulator (left term in Eq. 1), while the rest will be transferred by conduction to the PCM (

).

Similarly, part of the heat received by the PCM will be stored, while the rest will be transferred to the surrounding fluid by convection or absorbed due to the phase change in the PCM (i.e. latent heat). The problem resolution assumptions are the following: (i)

Super-cooling phenomenon is not considered.

(ii) The model does not take into account the natural convection in the PCM. For simplicity, only the heat transfer by conduction is considered. A detailed study taking into account convection as well could be the subject of a subsequent project. (iii) PCM physical properties are assumed to be constant. (iv) The temperature is assumed to be uniform in the cell. (v) The PCM melts at a constant temperature. (vi) Contact thermal resistance between the cell and the PCM is not taken into account (difficult to estimate when the PCM is in the solid state and almost zero in the liquid state). Based on energy balance equations and heat transfer laws, the following modelling equations can be deduced:

Inside the cell

Eq. 1

Inside the PCM [29]

Where

Eq. 2

denotes the total enthalpy (Eq. 3):

)

Eq. 3 Eq. 4

Eq. 5

According to the hypothesis (ii) and by neglecting the velocity inside the PCM i.e.

, (Eq. 2) can be reduced to (Eq. 6).

Eq. 6

By manipulating the PCM balance equation (Eq. 2) and (Eq. 4) the energy balance inside the PCM becomes:

Inside the PCM [18]

Eq. 7

The coupled problem (Eq. 7 and Eq. 1) was solved based on finite volume method combined with the Voller algorithm (for the liquid fraction determination), using an implicit discretization. More details about the solution are included in the appendix. The MATLAB code was validated numerically by comparison with 2D simulation results using Fluent software that is assumed to be reliable and robust. 2.2.2. 3D modelling: PCM integration Different options are available for PCM integration. A 3D modelling study was conducted to investigate what was the best integration architecture that could allow better heat rejection from the battery and maximize the PCM amount in order to prevent battery thermal runaway and non-uniform temperature distribution. To do that, three design schemes were simulated:

a)

b)

c)

d)

Figure 2: Representation of the three investigated designs. The high thermal conductivity material is indicated with blue color. (a) design with thin layer and plates. (b) design with only the upper plate (only 1/16 of the cell is represented). (c) design with fins. PCM and the cell are not represented (transparent). (d) mesh of the design ‘b’ o

In the first design, the cell was surrounded by a thin layer of a material with high thermal conductivity which was in contact with two plates of the same material. The latter permitted both the containment of the PCM and the heat removal by ventilation (Figure 2a). There is no direct contact between the PCM and the cell (Figure 2a).

o

In the second design, the cell was in direct contact with the PCM. There was no thin layer surrounding the cell (Figure 2b).

o

The third design was similar to the first one but some fins were added as well (Figure 2c). This design would make possible to know if fins could enhance the heat exchange efficiency from the PCM to the environment (air).

The three configurations were meshed and designed using Gambit software and solved using Fluent software (Volume of Fluid method was applied to solve the melting and solidification problem). A sensitivity analysis was conducted towards the time step and mesh size. A time step of 1s and a number of 20726 elements were the highest and lowest values respectively to guarantee time step and grid independence. 2.3. Modelling results

In this section, results of the parametric study regarding the melting temperature and the thermal conductivity and the system design are presented. Each study, including the experimental one, has been conducted independently from the others. Conclusions and findings are, however, general (independent from the considered PCM) and valid for future development of a complete full scale prototype of the PCM-based BTMS using the concept proposed in this study. 2.3.1. 1D modelling 2.3.1.1. Preliminary results In the following, temperature profiles were identified in the cell (heat source), unless otherwise stated. It is also important to note that it was considered that the battery does not discharge, this allows us to see the complete profile of the temperature even after the complete melting of the PCM. In reality, batteries generally are discharged after 90 minutes.

a)

b)

c) Figure 3: (a) Temperature evolution inside the cell surrounded by the PCM. (b) Temperature evolution inside the cell surrounded by the PCM showing the effect of the melting temperature on the phase change duration of the PCM. (c) Temperature evolution inside the cell surrounded by the PCM showing the effect of the thermal conductivity on the cell temperature stabilization. To show the effect of adding the PCM around the cell battery, the developed BTMS 1D model was used to plot the temperature evolution inside the cell surrounded by the PCM (Figure 3a). The power of the energy source is equal to 1.43 W (

), PCM and

cell characteristics are given in Table 1. The natural convection heat transfer coefficient of the surrounding air is supposed to be equal to 5 W.m-2.K-1. When the PCM is in its solid phase the cell temperature increases almost linearly by 12°C during 1150s. After that the cell temperature achieves the melting temperature of the PCM (35°C), as a consequence, a large part of heat released by the cell is absorbed by the PCM. As the melting occurs at constant temperature, during this phase the cell temperature remains below 40°C for almost 3000s. Figure 3a also shows that when the PCM is completely melted, the cell temperature starts

another increasing process but with merely less important rate. This could be explained by the fact that the difference between the PCM temperature and the ambient air is increasing; thus, the heat transfer to the environment is much higher than that of the beginning of the simulated experiment. [°C]

[W.m-1.K-1]

[kg.m-3]

[J.kg-1]

[J.K-1.kg-1]

PCM [30] [31]

35

0.2

820

157,000

2,100

Cell

--

--

2,047

--

1,075

Table 1: PCM and cell characteristics for 1D modelling This preliminary study shows that the PCM is useful only before it completely changes its phase. The authors of this paper investigated the regeneration of the PCM (transformation from the liquid to solid phase in order to increase the operation time with the PCM having the desired phase) using a forced ventilation system. This solution would be the most efficient one, because it would be possible to use the PCM in its desired phase as much as needed, but the results show that a high power (554W) is needed to make this transformation in the desired duration. As a result, the ventilation system would no longer be simple and convenient for BTMS of EVs. Thus, it is very important to prolong the melting process last as long as possible and find a solution to enhance the heat transfer to the environment in case of a complete phase change. In fact, because the PCM amount is limited (for any system), the complete melting/solidification of the PCM will certainly happen during the long charging/discharging periods. The evolution of melting (or solidification) front is governed by the following equation (known as Stephan condition [32]):

Eq. 8

where is the interface position. (Eq. 8) states that the phase change duration (image of the phase shift as the thickness of the PCM layer is constant) depends particularly on the thermal conductivity, latent heat and the mass of the PCM and its melting temperature. Knowing this information, it is necessary to make a parametric study to determine the characteristics of the PCM which can result in the best system. 2.3.1.2. Effect of PCM melting temperature In this section, the effect of the melting temperature where the (Eq. 9) is defined, is presented. Using the 1D model, a simulation of the cell temperature for different cases was performed. Each case corresponds to the same PCM but with different melting temperature. The results are shown in Figure 3b and Figure 3c. A security temperature in this case was chosen equal to 48 °C (the choice of the reference value is arbitrary). Results (Figure 3b) show that the higher the melting temperature, the longer the phase change duration (increase of about 2000s for

= 43 °C compared to

= 27 °C). This phenomenon is due to the fact that the heat transfer from the PCM to the air is high

(compared to cases where the melting temperature is low) when the PCM temperature is high (PCM receives less heat to change its phase). For the case of the PCM with the lowest melting temperature, the cell temperature achieves the security limit much

sooner than the other cases. PCM selection should, as a result, take into account the melting temperature regarding the maximum allowed temperature of the battery pack. 2.3.1.3. Effect of PCM thermal conductivity A parametric study similar to the previous one but regarding the thermal conductivity was investigated. Results presented in Figure 3c clearly show the benefits of an increased value of the thermal conductivity. A PCM with high thermal conductivity is able to reject much heat from the cell to the environment. As a result, the temperature of the cell is flattened during the phase change process, this phenomenon stops the temperature rise for a long period of time (as long as the PCM is not completely melted). For instance, the system with

= 16 W.m-1.K-1 is the last one which reaches the temperature limit (48 °C).

2.3.1.4. Improved PCM-based BTMS In Figure 4a, the temperature inside the cell falls in two systems, the first one with the reference PCM (Table 1) and the second one using a PCM whose thermal conductivity and melting temperature were improved. It is clearly noticeable that the improvement of physical properties of the PCM helps mitigate the increase in temperature more effectively. Indeed, the cell temperature (red curve) stabilizes from 2100s to 6100 approximately, this can significantly delay the temperature rise and thus protect the cell. Compared to the reference system, the cell temperature of the improved PCM reaches the security limit more than 1000s later.

a)

b) Figure 4: Temperature evolution inside the cell surrounded by the PCM. (a) comparison between the reference (the commercial PCM) in red line and the improved PCM in dark line. (b) comparison between a ventilated and a non-ventilated system.

Finally, and as expected, by using some forced ventilation the system performance is further improved. The corresponding results are presented in Figure 4b. In fact, the melting process of the ventilated PCM (convection heat transfer coefficient equals to 15 W.m-2.K-1) takes longer; thus, the cell temperature increases with a lower rate compared to the system with natural ventilation (convection heat transfer coefficient equals to 5 W.m-2.K-1). The various simulations have shown the usefulness of increasing the thermal conductivity of the PCM, selecting a PCM with a melting temperature closer to the security limit temperature (allowed by the battery) and coupling the PCM-based BTMS to a system of ventilation which should be properly sized. The purpose of these three measures is to limit the supply of heat, in time, to the PCM and thus extend its phase change duration and increase the heat transfer to the environment. The results of this 1D modelling investigation are interesting and promising, however, one can ask about the ease of finding a compromise between thermal conductivity, the melting temperature and the power of ventilation that would allow having the desired results. Further research on the possibility of developing tailor-made PCM and experimental investigations of various types of PCM are required to address this issue. 2.3.2. 3D modelling Due to the symmetricity of the geometry and by considering uniform heat transfer from the battery to the environment, as well as taking into account the study hypothesis presented earlier, it is possible to focus on modelling only 1/16 of a cell (1/4 x 1/2 x 1/2 half of the cell-) and extrapolate the results to the complete battery pack. For this reason, the developed 3D model focuses only on a small part of one cell. In this section, a composite PCM has been considered with sufficiently high thermal conductivity (16.6 W.m-1.K-1) [6] (see Table 2). For all the designs, simulations have been carried out with a heat transfer coefficient equals to 15 W.m-2.K-1 for convection from the plates to the surrounding air. The power of the energy source is equal to 1.43 W. The high thermal conductivity material considered for this investigation is aluminum (see Figures 2). [°C] PCM

43

[W.m-1.K-1]

[kg.m-3]

16.6

866

[J.kg-1] 181,000

[J.K-1.kg-1] 1,980

Table 2: PCM characteristics for 3D modelling Results of the 3D modelling are presented in Figure 5. The temperature evolutions have been plotted for each design concept for two locations (in the center of the cell and in the PCM at the corner). Note that the solution allowing the stabilization of the temperature as long as possible (red curves) is the design without the thin aluminum (Al) layer and without fins (only the two ventilated plates and PCM are integrated). The addition of fins or a thin layer of Al limits the amount of the integrated PCM, reducing the heat storage capacity. For the design considered to be the best, the cell temperature reaches the security limit more than 1000s later compared to the two other designs. The temperature contours, in the PCM, the cell, and the plates for the best design are given in Figures 5b. Results show that the system provides a uniform temperature distribution in the battery cell, the maximum temperature difference is less than 2°C. This is very important for maintaining the state of health of battery cells.

Corner

Center

Cell

a)

b) Figure 5: (a) Comparison between temperature evolutions for each design concept. (b) Temperature contour in the 1/16 of the studied system (at t=8481s) for the best design without the aluminum thin layer and fins 2.4. Modelling outcomes 1D and 3D modelling results showed that it is theoretically possible to improve the performance of a BTMS using a well selected PCM, with high thermal conductivity and well selected melting temperature regarding the battery limit temperature, combined with a simple forced ventilation system and, all these elements, are integrated in a specific design architecture (see Figure 2). All these findings have been used to develop an innovative BTMS. Figure 6 represents the suggested battery pack architecture with the integrated PCM. This BTMS concept is based on the best design determined by simulations in previous sections. The improved PCM is surrounding the cells. The PCM and the cells are confined between two plates of a high thermal conductivity material (Al in this study) which is ventilated using a low power fan.

Figure 6: The suggested design of a complete battery pack with BTMS with upper and lower plates under ventilation Besides the improved design, the second key parameter of the proposed novel PCM-based BTMS system is the way the PCM thermal conductivity was improved. An innovative solution is the one that is able to enhance the thermal performance of the PCM at the lowest cost. To do that, a simple but effective method of integrating a square grid array of copper, stacked around the cell and filled with the PCM in its liquid phase, was identified. Due to the symmetricity and for simplification reasons only one cell has been considered for the experimental study. The cell is actually a cylindrical element heated using Joule effect (fake cell) [11]. 3.

Experimental study

3.1. Small-size BTMS prototype development The complete system that surrounds the fake cell is shown in Figure 7. The transparent Plexiglas support was insulated during the experiments in order to reproduce the real symmetric conditions. The Plexiglas support was first filled with copper dutch weave then the fake cell was inserted between them. Thereafter, the remaining gap was filled with PCM and finally the upper plate was fixed. A commercial fan was used to horizontally ventilate the two aluminum plates.

O-ring sealing O-ring sealing

Upper plate in Al

Support in Plexiglas TC

Copper

Lower plate in Al

to atmospheric pressure

Figure 7: Pictures of the experimental setup highlighting the developed system, thermal insulation and the support so that the ventilation by the fan can reach the upper and bottom plates. Six experiments have been conducted using different heat source powers. The investigated case studies are given in Table 3.

Natural convection

= 2.45W Improved PCM

♣

= 6.10W ♣

Forced convection

= 14W ♣

= 2.45W ♣

= 6.10W

= 14W

♣

♣

Basic PCM

Table 3: The six case studies investigated 3.2. Experimental results and discussion Using 4 thermocouples inserted through the upper plate (see small holes in Figure 7), temperatures inside the PCM next to the fake cell (at the same distance from the cell but at different depths: 5cm; 2cm; 2.6cm; 2cm for TC4, TC3, TC2, TC1 respectively) have been measured for different configurations: with natural or forced convection, with or without the copper grid and with different thermal power source (Table 3). The PCM considered in this study was Paraffin RT35 whose melting point is equal to 35°C and its characteristics are given in Table 1; further characteristics of the considered PCM are given in Ref. [30] and [31]. The ventilation convective heat transfer coefficient was estimated using the correlation of forced convection around a flat plate (Eq. 9 and Eq. 10) to be equal to 16 W.m-2.K-1.

Eq. 9

Eq. 10

To show the effect of the suggested method for PCM thermal conductivity enhancement, two experiments have been conducted. The first one consisted of a PCM with no conductivity enhancement (no copper added) and the second one with the copper grid inserted. Figure 8a shows the temperature evolution inside the PCM at two different positions (TC1 and TC2) for the two experiments. As expected, the presence of the copper grid has greatly influenced the evolution of the temperature. Indeed, with a design incorporating copper and PCM, temperatures are merely 10°C lower than the design incorporating only the PCM. Results also show that the copper helps improving the temperature uniformity. In fact, as it can be noticed in Figure 8a, the temperature difference between TC1 and TC2 was much higher during the case where the PCM was used without the copper grids. The proposed improvement method is efficient to extract heat from the cell and transfer it to the surrounding PCM (PCM horizontal sections and thus to its complete volume) and not only to the PCM which is in contact with the cell (usually resulting in an increase of the cell temperature due to the low thermal conductivity of the PCM).

a)

b)

c) Figure 8: (a) Temperature evolution measured with 2 TCs at 2 different positions. Influence of the inserted copper grids with high thermal heat source (14W). The dashed lines correspond to the case without conductivity enhancement (only PCM, no Copper) under natural ventilation. (b) Temperature evolution measured with 4 TCs at 4 different positions. Influence of ventilation with low heat source (2.45 W) in the presence of copper. The dashed lines correspond to the forced ventilation case. (c) Temperature evolution measured with 2 TC at 2 different positions inside the PCM for three different thermal source powers. Enhanced PCM under natural convection.

Results in Figure 8b show that, for a low thermal energy source, the forced ventilation was capable of maintaining the PCM temperature, around the cell, below 30°C for a long period of time until the energy cut off (at 8000s). Even before the PCM meets its melting point, the ventilation was able to keep a temperature difference more than 5°C compared to naturel convection. Since in reality the melting occurs in a range (2°C width for the considered PCM) the temperature evolution is no longer perfectly flat as shown in numerical modelling results. The heat generated inside the fake cell is uniform but due to the flat plate in contact with air, a non-uniformity of temperature in the system can be noticed (depending on the depth of the TC). Temperature in the center region is the highest, this is why PCM at the center of the system starts melting first. Thanks to the PCM, the temperature nonuniformity is limited to approximately 2°C. The developed enhanced PCM limitations are revealed for high thermal energy sources. In fact, as it can be noticed in Figure 8c, the PCM achieves the considered security limit temperature (47°C) much earlier compared to the low heat source cases. These results show the importance of implementation of electronic systems capable of cutting off the electric current in order to prevent thermal runaway when the generated heat inside the battery is out of the PCM prevention capabilities. While the present paper suggests an efficient way to integrate PCM within a BTMS, it is important to mention that the thermal resistance present between battery body and copper foam/PCM is still a challenge that should be solved to improve farther the system performance. One potential method to reduce the thermal contact resistance maybe the integration of a system to maintain a high pressure within the battery to increase the contact area and thus heat exchange. This technical problem should be a priority for future research studies. 4.

Conclusion and perspectives

A new way to improve the performance of PCM and its integration in BTMS through the investigation of the effect of PCM melting temperature and enhancement of its thermal conductivity were studied in this research paper. Two numerical models, 1D and 3D modelling, have been exploited to define a new system design for an efficient PCM integration in battery packs and show the importance of enhancing the heat thermal transfer inside the PCM and its selection. Results from the numerical studies have been subsequently used to develop a small scale prototype of a PCM-based BTMS. The novelty of the proposed system concerns a new way of enhancing the thermal heat transfer in the PCM by using copper grids. The conducted experiments show very promising results. In fact, the new system is definitely more efficient compared to PCM-based BTMS without any enhancement regarding the PCM heat transfer capabilities. The suggested system was able to limit the temperature raise close to the cell by more than 5°C and make temperature uniform around the cell compared to the reference solution (without enhancement). For further improvement and system optimization, conducting the following research investigations are of great interest: 1.

Development of a real size prototype (battery back with real cells) using the proposed new solution and identification of a 3D model of the battery pack based on experimental data.

2.

Development of a test bench to measure and quantify the effect of contact thermal resistance between the battery cells and the copper foam/PCM.

3.

Methods to reduce the thermal resistance between the battery cells and the copper foam/PCM.

4.

Determination of the key parameters (PCM selection and amount, high thermal conductivity metal grid dimensions and shape, upper and lower plate thickness, air flow rate, etc.) using advanced optimization algorithms able to handle continuous and integer variables and the 3D model.

5.

Investigation of the security aspects and integration in EVs of such a new solution.

Acknowledgment This study has been supported and funded by the French Alternative Energies and Atomic Energy Commission (CEA). References

[1] S. Dechert, 15 3 2015. [Online]. Available: http://cleantechnica.com/2015/03/15/big-deal-economyenergy-co2-decoupling/. [Accessed 25 7 2016]. [2] K. Zaghib, A. Mauger and C. Julien, "Rechargeable lithium batteries for energy storage in smart grids," Rechargeable Lithium Batteries From Fundamentals to Applications, p. 319–351, 2015. [3] P. Mesarića and S. Krajcarb, "Home demand side management integrated with electric vehicles and renewable energy sources," Energy and Buildings, vol. 108, pp. 1-9, 2015. [4] A. Dinger, R. Martin, X. Mosquet, M. Rabl, D. Rizoulis, M. Russo and G. Sticher, "Batteries for Electric Cars: Challenges, Opportunities, and the Outlook to 2020," BCG, 2010. [5] P. Ramadass, R. Haran, R. White and B. Popov, "Capacity fade of Sony18650 cells cycled at elevated temperatures, part II. Capacity fade analysis," Journal of Power Sources, vol. 112 , pp. 614-620, 2002. [6] R. Sabbah, R. Kizilel, J. Selman and S. Al-Hallaj, "Active (air-cooled) vs. passive (phase change material) thermal management of high power lithium-ion packs," Journal of Power Sources, vol. 182, p. 630–638, 2008. [7] S. Al Hallaj and J. R. Selman, "A Novel Thermal Management System for Electric Vehicule Batteries Using Phase-Change Material," Journal of the Electrochemical Society, vol. 147, pp. 3231-3236, 2000. [8] Z. ho Rao and S. Wang, "A review of power battery thermal energy management," Renewable and Sustainable Energy Reviews, vol. 15, pp. 4554-4571, 2011. [9] L. Ziye, Z. Zhengguo, S. Guoquan, F. Xiaoming, W. Lei, G. Xuenong, F. Yutang, X. Tao, W. Shuangfeng and L. Xiaohong, "Review on thermal management systems using phase change materials for electronic components, Li-ion batteries and photovoltaic modules," Renewable and Sustainable Energy Reviews, vol. 31, p. 427–438, 2014. [10] Z. Qu, W. Li and W. Tao, "Numerical model of the passive thermal management system for highpower lithium ion battery by using porous metal foam saturated with phase change material," International Journal of Hydrogen Energy, vol. 39, p. 3904–3913, 2014. [11] C.-V. Hémery, F. Pra, J.-F. Robina and P. Marty, "Experimental performances of a battery thermal management system using a phase change material," Journal of Power Sources, vol. 270, p. 349– 358, 2014.

[12] Q. Wang, Z. Rao, Y. Huo and S. Wang, "Thermal performance of phase change material/oscillating heat pipe-based battery thermal management system," International Journal of Thermal Sciences, vol. 102, p. 9–16, 2016. [13] N. O. Moraga, J. P. Xamán and R. H. Araya, "Cooling Li-ion batteries of racing solar car by using multiple phase change materials," Applied Thermal Engineering, vol. 108, p. 1041–1054, 2016. [14] S. A. Khateeb, M. M. J. R. Farid and S. Al-Hallaj, "Design and simulation of a lithium-ion battery with a phase change material thermal management system for an electric scooter," Journal of Power Sources, vol. 128 , pp. 292-307, 2004. [15] S. Khateeb, S. Amiruddin, M. Farid, J. Selman and S. Al-Hallaj, "Thermal management of Li-ion battery with phase change material for electric scooters: experimental validation," Journal of Power Sources, vol. 142 , pp. 345-353, 2005. [16] R. Kizilel, A. Lateef, R. Sabbah, M. Farid, J. Selman and S. Al-Hallaj, "Passive control of temperature excursion and uniformity in high-energy Li-ion battery packs at high current and ambient temperature," Journal of Power Sources, vol. 183, p. 370–375, 2008. [17] Y. Lv, X. Yang, X. Li, G. Zhang, Z. Wang and C. Yang, "Experimental study on a novel battery thermal management technology based on low density polyethylene-enhanced composite phase change materials coupled with low fins," Applied Energy, vol. 178, p. 376–382, 2016. [18] G. Karimi, M. Azizi and A. Babapoor, "Experimental study of a cylindrical lithium ion battery thermal management using phase change material composites," Journal of Energy Storage, 2016. [19] F. Samimi, A. Babapoor, M. Azizi and G. Karimi, "Thermal management analysis of a Li-ion battery cell using phase change material loaded with carbon fibers," Energy, vol. 96, p. 355–371, 2016. [20] Z. Wang, Z. Zhang, L. Jia and L. Yang, "Paraffin and paraffin/aluminum foam composite phase change material heat storage experimental study based on thermal management of Li-ion battery," Applied Thermal Engineering, vol. 78, pp. 428-436, 2015. [21] W. Li, Z. Qu, Y. He and Y. Tao, "Experimental study of a passive thermal management system for high-powered lithium ion batteries using porous metal foam saturated with phase change materials," Journal of Power Sources, vol. 255 , pp. 9-15, 2014. [22] C. Lin, S. Xu, G. Chang and J. Liu, "Experiment and simulation of a LiFePO4 battery pack with a passive thermal management system using composite phase change material and graphite sheets," Journal of Power Sources, vol. 275, pp. 742-749, 2015. [23] G. Jiang, J. Huang, Y. Fu, M. Cao and M. Liu, "Thermal optimization of composite phase change material/expanded graphite for Li-ion battery thermal management," Applied Thermal Engineering,

vol. 108, p. 1119–1125, 2016. [24] Z. Ling, F. Wang, X. Fang, X. Gao and Z. Zhang, "A hybrid thermal management system for lithium ion batteries combining phase change materials with forced-air cooling," Applied energy, vol. 148, pp. 403-409, 2015. [25] Z. Ling, J. Chen, X. Fang, Z. Zhang, T. Xu, X. Gao and S. Wang, "Experimental and numerical investigation of the application of phase change materials in a simulative power batteries thermal management system," Applied energy, vol. 121, p. 104–113, 2014. [26] Z. Rao, Q. Wang and C. Huang, "Investigation of the thermal performance of phase change material/mini-channel coupled battery thermal management system," Applied Energy, vol. 164, p. 659–669, 2016. [27] M. Muhieddine, E. Carnot and R. March, "Various Approches for Solving Problems in Heat Conduction with Phase Change," International Journal on Finite Volumes, vol. 6, pp. 66-85, 2009. [28] V. Voller, "Fast implicit finite-difference method for the analysis of phase change problems," Numerical Heat Transfer, Part B, vol. 17, pp. 155-169, 1990. [29] J. Crank, Free and Moving Boundary Problems, Oxford: Clarendon Press, 1984. [30] A. Caron-Soupart, J.-F. Fourmigué, P. Marty and R. Couturier, "Performance analysis of thermal energy storage systems using phase change material," Applied Thermal Engineering, vol. 98, pp. 1286-1296, 2016. [31] M. Longeon, A. Soupart, J.-F. Fourmigué, A. Bruch and P. Marty, "Experimental and numerical study of annular PCM storage in the presence of natural convection," Applied Energy, vol. 112, pp. 175184, 2013. [32] S. Liu, Y. Li and Y. Zhang, "Mathematical solutions and numerical models employed for the investigations of PCMs‫ ׳‬phase transformations," Renewable and Sustainable Energy Reviews, vol. 33, pp. 659-674, 2014.

Appendix The finite volume method consists in integrating, on elementary volumes, the equation of the energy balance written in integral form. The computational domain is divided into

meshes of center

and a variable volume

. Each mesh has a size

. The half-integer indices denote the interfaces of the mesh with the neighboring meshes (see Figure A-1).

Figure A-1: Finite volumes elements. The time is discretized in intervals of constant time step formula [27]. The functions s,

or

, the choice of the time step has been made while respecting Edouard's

are assumed to be uniform in each mesh but depend on time.

Each mesh exchanges heat with its two neighbors

and

and stores a part of it.

Two halves a mesh have been added respectively to the extreme surfaces (from cell side and ambiance side) of the computational domain in order to be able to easily express the boundary conditions at the mesh input and output (Cell temperature temperature

and air

).

The discretization of the equations was carried out with an implicit method better suited to our problem. We consider that the heat flux transferred to the PCM is a conduction flux, the discretization of (Eq. 7) using the finite volume method gives:

Eq. A-1

Eq. A-2

Eq. A-3

Eq. A-4

Where

,

,

and

are respectively the delimiting surface at

surfaces and the sensible enthalpy for mesh # and at time step

and

the volume between these two

.

In Table A-1 are compiled the expressions of the constants that appear in the discretized balance equation for boundary conditions. The description of the elementary volumes and surfaces is presented in Figure A-1. Coefficients

Expression or value 0

0 The following matrix equation is solved each time step to compute

for each mesh: Eq. A-5

Where:

 hs1n 1   b1  n 1    hs2   a2   H   hs3n 1  ; M         n 1  0   hs N x  While, the liquid fraction i.e. conditions.

c1 b2 a3

c2 b3 

  a Nx

 hs1n   n   0      n  hs2     ; Q   hs n    3        c N x 1    n    n bN x   hs N x    

is calculated using the Voller method.

 n and  n

f1n  f1n 1   f 2n  f 2n 1  f 3n  f 3n 1     f Nnx  f Nnx1 

are determined as a function of the boundary

HIGHLIGHTS



Investigations show the importance of selecting a PCM with high melting temperature



A solution to enhance heat transfer inside the PCM by copper dutch weave was developed



PCM-based BTMS reduces temperature rise more than 5°C and improves its distribution



Simplified 1D and 3D models of a PCM-based BTMS were developed