Please submit your abstract online at ICAS website ...

Regional Conference on Mechanical and Aerospace Technology Bali, February 9 – 10, 2010

Thermal Engineering Performance Evaluation of a Polymer Electrolyte Membrane Fuel Cell Stack at Partial Load W.A.Najmi W.M.*, Rahim A. and M. Fairuz. R. Alternative Energy Research Centre, Faculty of Mechanical Engineering, Universiti Teknologi MARA, 40450 Selangor, Malaysia. Email : [email protected] Keywords: PEM fuel cells, thermal engineering, temperature profile, cooling, heat generation . Abstract : Thermal effects are critical factors on the performance and life of a Polymer Electrolyte Membrane (PEM) fuel cell system. Heat generation in a fuel cell stack is largely influenced by the exothermic electrochemical reactions at the cathode. Due to nonuniformity of reactions over the MEA area, heat generation and temperature distribution over the stack is also theoretically non-uniform. This work attempts to profile the thermal performance of an industrial PEM fuel cell stack. The stack power generation and cooling rates at the heat exchanger were monitored under specific working conditions. The discussion focused on causes of the temperature gradient, heat generation, cooling effectiveness, and the quality of the heat exchanger. The analysis demonstrates a significant relationship between stack temperature and generated power to the heat exchanger performance and coolant conditions.

Keywords : PEM fuel cells, thermal engineering, temperature profile, cooling, heat generation

1. Introduction Polymer Electrolyte Membrane fuel cells have emerged as the most highly applied fuel cell system due to its capability of higher power density and faster start-up than other fuel cells. A PEM fuel cell directly converts the chemical energy within a fuel into electricity by the electrochemical reaction between a hydrogenbased fuel and oxygen from air. When hydrogen is used as the fuel source, only water and heat are formed as by-products, providing a much-sought alternative to clean energy generation in this modern but polluted world. PEM fuel cells operate at low temperatures of usually less than 100oC, allowing faster start-ups and immediate response to changes in the demand of power. Currently, the power range for PEM fuel cell is between 1 kW to 250 kW size systems. Thermal management efficiency dictates the operating conditions and output of a PEM fuel cell. The significance of thermal effects on the performance of PEM fuel cells was reviewed by Faghri and Guo [1] and Kandlikar and Lu [2]

where they agreed that significant technical challenges still exist in the aspect of thermal engineering. The heat generation is governed by inter-related factors such as thermal resistance, electrode properties, electrochemical reaction rates, reaction uniformity, humidity of reactants, as well as membrane permeability [3]-[5]. The energy conversion efficiency of a PEM fuel cell is 40% to 60% of the Lower Heating Value of the fuel [6]; thus, a PEM fuel cell with 50% efficiency generates an equal amount of electrical and thermal energy. The ability of the thermal engineering system to dissipate the generated heat effectively optimizes the power output and leads to longer cell life. Cooling failure leads to higher energy losses due to increase in membrane resistance to proton conductance [7], as well as membrane rupture that cause fuel leakage or crossover [8], critically causing combustion to occur within the cells. Zhang et.al [9] views the issue of thermal management for fuel cell systems consisting of two primary levels; first at the cell level to ensure proper membrane hydration, and secondly, at the system level to keep the stack from heating up.


The context of this paper are quantitative thermal analysis based on experimental works on a watercooled PEM fuel cell system, which is available at the Alternative Energy Research Center (AERC), Faculty of Mechanical Engineering, Universiti Teknologi MARA (UiTM) Malaysia. Data on dynamic temperature profile under certain operating conditions were obtained and compared with general models of heat generation and heat transfer of the stack and heat exchanger. Thermal and cooling effectiveness analysis was conducted to identify the practical operating relationships to the thermal characteristics of the system.

Equations 2 to 14 are applied accordingly for the analysis where the scope covers the relation of stack thermal conditions to output power, activeto-passive cooling contributions, and cooling system effectiveness. The electrical power output in Watts, (2)

For a single cell, the cell efficiency Vcell V  cell Vmax 1.482

Vstack, measured Vstack,rated



Vstack,measured 48V







(5)





Q a  ma .C p, a . Ta, e  Ta,i



(6)

The subscripts cw and a are for cooling water and air respectively, while subscripts e is for the exit state and i for the inlet state. The net heat in the stack related to the temperature difference over a certain time period, 

Q stack 

mstack.Cstack.Tstack t

(7)

Passive cooling over the exposed stack surfaces consists of free (natural) convection by the ambient surroundings as well as heat transfer by radiation. The free convection cooling effect is based on Qnc = h.Asurface (Tsurface-Tambient)

(8)

In this case, the surface areas involved are two vertical flat side surfaces, two vertical flat end surfaces, and one horizontal surface with heated surface facing upward. The Nusselt number correlations were calculated for each orientation and the respective free convection coefficient, h, and surface cooling is calculated.

(3) Cooling by radiation heat transfer is expressed by

Thus, the stack efficiency was evaluated using stack 



Qcw  mcw .C p,cw. Tcw,e  Tcw,i

(1)

In equation (1), I represents the stack current in Ampere (A) and ncell is the number of cells of the stack.

 cell 

Cooling water energy changes,



Stack heat generation is a theoretical heat quantity based on the conversion efficiency of the fuel cells. The theoretical maximum voltage for a single cell based on the Higher Heating Value of hydrogen is 1.482V when the products are all in liquid state. The stack for this work has a design rating of 45% efficiency at full load, and the theoretical heat generation for the stack can be estimated using Equation (1).

Pel = V x I

The calculation of active cooling rate is based on the energy property changes of the cooling water as it carries away the heat from the fuel cell stack.

Air energy changes,

2. Thermal Engineering Analysis

Qth  1.482 1   cell I .ncell

surrounding. Active cooling theoretically contributes at least 90% [12] of the total cooling effect. Passive cooling plays only a minor role, but the effects are more significant as the stack temperature increases.

(4)

The cooling mechanism of the fuel cell stack is categorized as active cooling and passive cooling. Active cooling is achieved by circulating cooling water internally and using an air-cooled heat exchanger (radiator) to dissipate the heat to the



Qr   .  As . . Ts 4  Tambient4



(9)

σ is the Stefan-Boltzmann constant, equals to 5.67 x 10-8, and ε is the surface emissivity. Thus, the passive cooling rate is the summation of free convection and radiative cooling over the stack surfaces, Qpassive = Σ Qnc + Qr

(10)


The total cooling effect is the summation of active and passive cooling rates. Σ Qcooling = Qactive + Qpassive

(11)

From equations (7) and (11), the total stack thermal power can be calculated. Pth = Qstack + Σ Qcooling

(12)

For the radiator, the analysis on the radiator effectiveness is evaluated by,

Table 1: PEM fuel cell system specifications Specification Power output rating Number of cells Operating temperature Cell size (bipolar plate) Cooling system





Q actual

Information / Details 3 kW at 48V 72 50oC 150 mm length, 240 mm height, 5 mm thick Water-cooled with heat exchanger

(13)



Q max

Table 2: Operating conditions PEM Fuel Cell stack

The maximum possible cooling rate in a heat exchanger is

Properties / parameters

Values



Q max  Cmin (Thot,in  Tcold ,in )

(14)

1.

Material

where Cmin is the smaller of Ch  mh C p , h and

2. 3. 4. 5. 6. 7.

Specific heat, C [13] Density [13] Stack volume Top surface area Side surface areas Thermal conductivity [14] Surface emissivity [15] Conversion efficiency Reactants





Cc  mc C p ,c .

3. Experimental Method The experiment was conducted using a PEM fuel cell system designed for Uninterrupted Power Supply (UPS). The system configuration and general specifications of the hardware are presented in Figure 1 and Table 1 respectively. The working fluids are hydrogen, reactant air, cooling water and cooling air. Table 2 summarizes the operating conditions of the experimental. The measurements were taken at 3 minutes intervals using a thermal scanner and Ktype thermocouples with data logger for local temperatures at 18 designated points of the stack, anemometer for air velocity and temperature, and a multi-meter for electrical power measurement at the resistant loader.

8. 9.

1. 2.

1. 2. 3.

1. 2. 3. 4. 5.

Carbon graphite 710 J/kg.K 2240 kg/m3 0.010557 m3 0.05775 m2 0.1848 m2 20 W/m.K 0.85 45%


Values

Hydrogen inlet pressure Air inlet pressure Cooling Water

1.5 bar 1 bar


Values

Operating pressure Specific heat, Cp [15] Mass flow rate Cooling Air

1 atm 4180 J/kg.K 0.0126 kg/s


Values

Inlet temperature Inlet pressure Specific heat, Cp [15] Mass flow rate Density [15]

≈ 26oC 1 atm 1007 J/kg.K 0.065274 kg/s 1.174 kg/m3


PUMP

BLOWER AIR STREAM

Twater,exit

Pair

PH2

COOLING WATER STREAM

PEM FC

COOLING AIR STREAM Tair,in

Tair,exit

HYDROGEN TANK H2 RECYCLE STREAM

RADIATOR PURGE Twater,in IONIZER

POWER MANAGEMENT UNIT

RESISTIVE LOADER

Figure 1. The fuel cell system schematic

4. Results and Discussion 4.1 Thermal Profile During the experiment, the electrical loader was approximately constant at 12A. It was observed that the voltage reading was between the range of 27.7V to 32.5V, giving a maximum power output of 440W at the early stages of the experiment. Comparing the electrical power output to the average stack temperature in Figure 2, it is noted that a sharp decline in electrical power was registered as the stack temperature rises. The designated operating temperature of the membrane was 50oC; however, the power decline was significant even as the stack temperature crosses 40oC and stabilizes at 300W as the temperature exceeds 50oC. The average power decline for the stack was calculated at 7.6 W/oC stack temperature increase. The results indicates a significant relationship on the effects of stack temperature to the power output, where the power decline is directly related to increased proton conductance resistance as the membrane dehydrates, as well as increased ohmic resistance of the electrodes as the temperature increases. Further investigation was conducted to identify the actual reasons behind the high stack temperatures encountered. Figure 3 indicates that the combined cooling effects of active and passive cooling was inadequate to effectively remove the

total heat within the stack, leading to a significant amount of net stack heat that increases the stack temperature. The cooling rate increases linearly at 73.3 W/min from the 7th to 22nd minute, and steady-state cooling was observed after the 22nd minute between 1100W to 1300W, causing a higher net stack heat buildup and consequently, a faster temperature rise at approximately 2.7oC per minute. The stagnant cooling capability at the later stages suggests a failure of the cooling system to provide an equal response to the cooling needs of the stack. Generally, the bulk of the thermal power of a fuel cell stack is governed by the rate of heat generation from the electrochemical reactions. However, the cooling system was designed to adequately handle the estimated thermal power generation, thus a study on the relationship between the net stack heat, stack temperature and cooling water exit temperature (Figure 4) was conducted to investigate the effects of external heat sources to the stack temperature. The graph points to the strong influence of cooling water temperature to the overall stack temperature. A thermal equilibrium condition between the stack and cooling water was identified from the 7th to 22nd minute, leading to the conclusion that the high stack temperatures obtained was actually caused through heating by the cooling water flow. From the 22nd minute onwards, the cooling water


was thermally saturated at 60oC, causing a larger net stack heat buildup from 380W to 720W, rapidly increasing the stack temperature by internal heating. The possible effects of membrane rupture, fuel crossover and combustion within the cells to the stack temperature rise are disregarded based on the amount of net stack heat buildup

which is acceptable with the rate of thermal power generated by electrochemical reactions. Thus, the cooling water influence on the stack temperature was identified as the main cause of stack temperature rise.

Figure 2. Relation of stack temperature to electrical power

Figure 3. The total stack thermal energy compared to combined cooling effects

Figure 4. Relation of stack heat and stack-to-cooling water temperature profile


Figure 5. Active and passive cooling contribution percentages Active and passive cooling distribution at variable stack power 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% 500

1000

1500

2000

2500

3000

3500

stack thermal power (Watts) % passive cooling

% active cooling

Figure 6.Theoretical optimum cooling contributions of passive and active cooling

4.2 Cooling Performance The cooling contribution percentage (or cooling effectiveness) is calculated based on the ratio of the rate of active or passive cooling to the required cooling rate of the stack for constant stack temperature. The active cooling rate is the heat removed by the cooling water, while passive cooling is the combination of natural convection and radiation heat transfer across the external stack surfaces. Figure 5 shows that the total cooling effect is less than 80% at all times, thus leading to stack heat increase and temperature rise. Steady-state active cooling is within the range of 54% to 67% cooling effectiveness, where optimum cooling occurs when the cooling water inlet temperature is less than 60oC. The active cooling contribution drops to 50% when the water inlet temperature rises above 60oC. Passive cooling contribution is significant only at high surface temperatures (larger difference with ambient). At stack temperatures less than 60oC,

the passive cooling effectiveness is only 11% to 12.5%. The effectiveness increases between 14% to 23% as the stack temperature increases from 60oC to 74oC. Ideally, the entire cooling system should be able to meet the necessary cooling load demand to maintain the stack at the desired operating temperature. The optimum cooling contributions of active and passive cooling at 50oC maximum stack temperature was theoretically analyzed and presented in Figure 6. In a 3 kW stack, the active cooling system is required to handle 90% of the cooling load, while the other 10% is handled by the passive cooling methods. Thus, the maximum quality (maximum cooling effectiveness achieved against required cooling effectiveness) of the active cooling in this study is approximately 75%.


Figure 7. Temperature profile at heat exchanger

Figure 8. Effectiveness of the heat exchanger at steady-state The heat exchanger fluid temperature (Figure 7) analysis was performed to pinpoint the technical reasons for an under-performing active cooling system. The exit condition of the cooling water and air from the heat exchanger is under thermal equilibrium for the whole process, meaning that the water inlet temperature to the stack is severely limited to the air cooling capability within the heat exchanger. The cooling water temperature into the stack increases at a rate of 1.2oC/min, or equivalent to 63.2 W/min of heat energy buildup. The effect of cooling water temperature to the stack temperature has been discussed earlier, thus ideally, the cooling water temperature throughout the stack should be limited to the required stack operating temperature, in this case 50oC. This leads to the conclusion that due to vaporization, the active cooling water flow rate circulating the stack was inadequate to meet the temperature increase limits. Analysis on the cooling air temperatures shows that the heat transfer rate is limited and constant at approximately 1000 W of cooling capability.

Ideally, the cooling water and air exit temperatures should be as low as possible without reducing the heat transfer rates between the two fluids. The air cooling capability is directly related to the flow rate of air through the heat exchanger; therefore, a larger airflow or powerful fan is needed. The fuel cell operating temperature can be effectively controlled by adjusting the cooling fan operation since the heat rejection at the radiator has been proven to be limited by the airflow rate. In another perspective, reducing the air exit temperature to at least 35oC would greatly enhance the cooling rate within the stack and limits the effect of cooling water temperature to the overall stack temperature. The heat exchanger effectiveness, as presented in Figure 8, shows how poorly the heat exchanger performs in meeting the cooling demands of the fuel cell system. The effectiveness is lower than 50%, and proven analytically to be the main reason for the thermal problems encountered by the fuel cell stack.


5. Conclusions Experimental analysis of a water-cooled PEM fuel cell system shows that the stack temperature can be largely influenced by the coolant. The stack temperature increased to unacceptable levels due to the influence from the cooling water temperature. To maintain the stack at the required operating temperature below 50oC, the generated heat needs to be adequately dissipated by an active cooling system operating higher than 90% cooling effectiveness, and the cooling water temperature at the inlet and outlets of the stack must also be controlled below 40oC at all times. With a registered effectiveness of less than 50%, unsuitable operating conditions of the heat exchanger was identified as the main cause for the thermal problem facing the stack, especially regarding the flow rates of both fluids.

References [1] A. Faghri and Z. Guo, “Challenges and opportunities of thermal management issues related to fuel cell technology and modeling”, International Journal of Heat Transfer and Mass Transfer, vol.48, no.1920, 2005. [2] S.G. Kandlikar and Z. Lu, “Thermal management issues in a PEMFC stack – a brief review of current status”, Applied Thermal Engineering, vol.29, 2009. [3] S. Pandiya, K. Jayakumar, N. Rajalakshmi and K.S. Dhathathreyan, “Thermal and electrical energy management in a PEMFC stack an analytical approach”, International Journal of Heat Transfer and Mass Transfer, vol.51, no.3-4, 2008. [4] S.K. Kamarudin, W.R.W.Daud, A. M. Som, A.W. Mohammad, S. Takriff and M.S. Masdar, “The conceptual design of a PEMFC system via simulation”, Chemical Engineering Journal, vol.103, 2004. [5] Y-J Sohn, G-G Park, T-H Yang, Y-G Yoon, W-Y Lee, S-D Yim and C-S Kim, “Operating characteristics of an air-cooling PEMFC for portable applications”, Journal of Power Sources, vol.145, 2005. [6] P.F. van den Oosterkamp, “Critical issues in heat transfer for fuel cell systems”, Energy Conversion and Management, vol.47, 2006. [7] R. Mosdale and S. Srinivasan, “Analysis of Performance and of Water and Thermal Management in Proton Exchange Membrane Fuel Cells”, Electrochemica Acta, vol.40, no.4, 1995. [8] F.C. Chen, Z. Gao, R.O. Loutfy and M. Hecht, “Analysis of optimal heat transfer in a PEM fuel cell cooling plate”, Fuel Cells, vol.3, 2003. [9] Y. Zhang, M. Ouyang, Q. Lu, J. Luo and X. Li, “A model predicting performance of proton exchange membrane fuel cell thermal systems”, Applied Thermal Engineering, vol.24, 2004. [10] C-Y Wen and G-W Huang, “Application of a thermally conductive pyrolytic graphite sheet


to thermal management of a PEM fuel cell”, Journal of Power Sources, vol.178, no.1, 2008. [11] S. Yu and D. Jung, “Thermal management strategy for a proton exchange membrane fuel cell system with a large active cell area”, Renewable Energy, vol.33, no.12, 2008. [12] W.A.N.W.Mohamed and R. Atan, “Optimizing air cooling capability of PEMFC through case-by-case cooling channel analysis”, Proceedings of the International Conference on Advances in Mechanical Engineering (ICAME), Malaysia 2009. [13] S. Begot and J.M. Kauffmann, “Estimation of internal fuel cell temperatures from surface temperature measurements”, Journal of Power Sources, vol.178, 2008. [14] L. Dummery, R. Glises, H. LouahliaGualous, J.M. Kaufmann, “Thermal management of a PEMFC stack by 3D nodal modeling”, Journal of Power Sources, vol.156, 2006. [15] Y. A. Cengel, Heat Transfer : A Practical Approach, Int. Edition. (McGraw Hills, New York, 2003).