- Email: [email protected]
Segmented cell approach for studying uniformity of current distribution in polymer electrolyte fuel cell operation
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Segmented cell approach for studying uniformity of current distribution in polymer electrolyte fuel cell operation Seung-Gon Kim, Min-Jin Kim, Young-Jun Sohn* Fuel Cell Laboratory, Korea Institute of Energy Research, 152, Gajeong-ro, Yuseong-gu, Daejeon 305-343, South Korea
article info
abstract
Article history:
In this study, the current distribution of a polymer electrolyte fuel cell (PEMFC) with a large
Received 13 February 2015
active area is investigated using a segmented cell system. A specially designed printed
Received in revised form
circuit board (PCB)-type segmented cell is applied to a single-cell PEMFC. By using the
6 May 2015
segmented system, the effects of clamping pressure uniformity between the components
Accepted 8 May 2015
of the PEMFC and the fuel injection direction are examined. The pressure uniformities of
Available online 10 June 2015
the two different types of endplates are measured using pressure indicating films (PIFs). A curved endplate is used to improve the pressure uniformity. The pressure uniformity is
Keywords:
found to significantly influence the current distribution in a PEMFC with a large active area.
Polymer electrolyte fuel cell
Two types of gas feeding modes, parallel and cross injection, are tested. The results show
Pressure uniformity
that the clamping pressure uniformity and gas feeding configuration affect the current
Segmented cell
distribution and overall PEMFC performance.
Pressure indicating film
Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Current distribution
Introduction Alternative power solutions such as wind power, solar power, and fuel cells have emerged owing to the growing demand for clean and renewable energy. Among these, polymer electrolyte membrane fuel cell (PEMFC) technology has attracted considerable attention as a mobile power solution owing to its relatively low operating temperature and suitable power range for vehicles [1]. However, the performance and durability of PEMFCs must be improved further before they can be commercialized. Generally, PEMFCs mainly suffer power loss owing to the contact resistance between the bipolar plates and the gas diffusion layers (GDLs) [2].
The contact resistance depends on the clamping pressure, gas pressure, current density, and temperature [3]. In Refs. [3], Ihonen et al. simultaneously measured the clamping pressure and contact resistance using a specially designed single cell. They validated the relationship between the contact resistance and other parameters through a comparison of in-situ and ex-situ measurements. Zhou et al. investigated the effect of the clamping force on the interfacial contact resistance and porosity of a GDL [4]. They suggested an optimal rib design for a gas channel for a reasonable combination of low interfacial contact resistance and good GDL porosity. Recently, pressure indicating films (PIFs) have been adapted to measure the pressure distribution. A PIF indicates
* Corresponding author. Tel.: þ82 42 860 3087; fax: þ82 42 860 3104. E-mail address: [email protected] (Y.-J. Sohn). http://dx.doi.org/10.1016/j.ijhydene.2015.05.055 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
11677
Fig. 1 e Segmented PEMFC. (a) Configuration of segmented PEMFC. (b) Photograph of PCB board.
applied pressure differences through color density variations, allowing the pressure distribution to be measured using software. Wen et al. investigated the fuel cell performance for different clamping torques and bolting configurations using PIFs. They found that a larger mean contact pressure leads to higher maximum power and uniformity of the contact pressure distribution. Furthermore, the ohmic resistance and mass transport limit current showed highly linear correlations with the mean contact pressure [5]. Yu et al. used a PIF to measure the pressure distribution in a single-cell PEMFC [6]. Instead of metals, they employed composite materials as endplates to increase the uniformity of the pressure distribution. They used composite endplates with a pre curvature created by thermal fabrication. These endplates red uced the stack weight without compromising the performance. In this study, PIF was used to measure the pressure distribution in a single-cell PEMFC. A curved endplate was used
to achieve high pressure uniformity. The effect of the pressure distribution was analyzed based on the current distribution. We adapted a printed circuit board (PCB)-type segmented cell to visualize the current distribution. Segmented cells were generally used to measure the current distributions [7,8]. Three techniques were used to realize a segmented cell system: PCB, resistor network, and Hall effect sensors [9]. Cleghorn et al. [7] first adapted the PCB-based measuring technique, in which a segmented flow field and a current collector were used. Wieser et al. [10] first introduced the Hall effect sensor for indirect current sensing. They investigated the relation between the uniformity of the pressure and the current distribution using a PCB-type segmented cell. As mass transport is one of the most significant phenomena in PEMFCs owing to the electrochemical reaction, many studies have investigated efficient flow configurations through numerical as well as experimental methods [11]. However, only a
11678
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
few studies have investigated the gas feeding modes. Ge and Yi investigated the influence of the gas feeding mode in a PEMFC fuel cell with an active area of 140 cm2 [12]. They compared the counter flow and the co-flow in terms of the membrane humidification and current density distribution. Scholta et al. studied the PEMFC performance for different gas feeding modes
of the co-flow, counter flow, and cross flow [13]. Recently, Sierra et al. classified the gas feeding configuration into four modes: cross flow, nonsymmetrical flow, similar flow, and counter flow [14]. They reported that the local current distribution in the catalyst layer can be improved by changing the gas feeding mode. In this study, the current distributions with different gas
Fig. 2 e Experimental setup of segmented PEMFC. (a) Schematic of experimental setup. (b) Photograph of experimental setup.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
feeding modes are obtained using a segmented cell system. As a uniform current distribution leads to a uniform temperature and liquid water production [11], better performance and durability can be achieved with a uniform pressure distribution and optimized gas feeding configuration.
Experimental method and apparatus Experimental setup Fig. 1(a) shows the configuration of a single-cell PEMFC with a PCB-type segmented cell. A segmented PCB with 112 (7 16) segments was installed between a gas flow channel and an endplate. The size of each segment was 1.79 cm 1.79 cm. Metal-build gas flow channels with a large active area of 360 cm2 were employed. The PCB collects the electric current generated by the PEMFC. Fig. 1(b) shows a photograph of the segmented PCB used in this experiment. Fig. 2(a) shows a schematic diagram of the experimental setup. The local current from the PEMFC passes through each segmented cell of the PCB. The current is distributed to the current sensors (AMETES CS2.5A-02, SENIS, Switzerland).
11679
These sensors are connected in the unipolar operation mode (0e5 V), corresponding to a current of 0e2.5 A (0e780 mA/ cm2) for each sensor. Therefore, the maximum current density was limited to 700 mA/cm2 for ensuring the safety of the system. The distributed current was collected to the electric load (PLZ664WA, KIKUSUI, Japan). Owing to the electrochemical characteristics of a single-cell PEMFC with a large active area, the electric load needs to be operated under low voltage (below 1 V) and high current (of the order of hundreds of amperes). Fig. 2(b) shows a photograph of the experimental setup.
Segmented cell calibration Although all segments are installed in the PCB, each electric path from the segmented cell surface to the data acquisition (DAQ) system varies. To compensate the ohmic difference between the PCB and the DAQ systems, a calibration procedure was performed using a calibration unit, as shown in Fig. 3(a). The calibration process was performed as follows. Fig. 3(b) shows a flowchart of the calibration process.
Fig. 3 e Current calibration method. (a) Segmented PCB calibration unit. (b) Flowchart of segmented PCB calibration. (c) Current calibration using four-point least squares method.
11680
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Fig. 4 e Pressure distribution of plain and curved endplates.
1. Charge the reference currents to the first segmented cell 2. From the measured values, formulate a transposing equation using linear squares fit. 3. Move to the next segmented cell. 4. Repeat steps 1e3 for every segmented cell until the last segmented cell. 5. Apply every transposing equation during current acquisition in the experiment. In this experiment, by applying the accurate current induced by the DC power supply as a reference value, the measured value can be corrected to the calibrated value using a least squares method. Four different current values (0, 200, 400, and 600 mA) were charged to a single segmented cell as reference currents. Fig. 3(c) shows the typical calibration result obtained using the least squares method. A solid line shows the current actually charged in a segmented cell by the DC power supply. However, the value measured by the current sensor and data acquisition system differ from the actual value. These gaps between the actual and the measured values increase at higher current density. After the calibration process described above, the calibrated values approach the actual current values.
Results and discussion Pressure distribution Two types of endplates were employed to compare the pressure distributions, which were of the plain and the curved type. Both endplates were made of 30-mm-thick epoxy glass. Fourteen M10 bolts were used for the combination of endplates. All bolts were fastened with the same torque of 110 kgf$cm. In the case of a curved endplate, the vertex depth was 0.5 mm. Fig. 4 shows the pressure distribution using a PIF. A
Table 1 e Pressure distributions according to the shape of endplate. End-plates shape
Plain end-plate
Curved end-plate
Pressed Area [mm2] Ave. Pressure [Mpa] Max. Pressure [Mpa] Load [N]
27761 0.77 3.06 21321
36416 0.96 3.06 34970
22 C, 46%RH, Continuous pressing mode.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
“Super Low” PIF (FUJIFILM, Japan) was used during the experiment at pressures of 0.5e2.5 MPa [11]. Quantitative pressure information was obtained using an FPD-8010E pressure distribution mapping system (FUJIFILM, Japan). Table 1 summarizes the experimental data. As shown in Table 1, the total pressed area and average pressure of the curved endplate is much larger than that of the plain endplate, although the maximum pressures have the same value. This result indicates that the curved endplate efficiently pressed the single-cell PEMFC.
Curvature effect of endplate on current distribution Fig. 5 shows the typical current distribution results of a singlecell PEMFC. The center region of the curved endplate shows higher activation status than the plain endplate. Owing to these activation differences, the cell voltage of the curved endplate is higher at the same current density than the plain endplate. Fig. 6(a) shows the overall current density curve of the two different types of endplates. As shown in the graph, the current density of the plain endplate rapidly decreased in the high current density region. This was mainly attributed to
11681
increased ohmic contact resistance owing to the pressure distribution. As the effect of the ohmic resistance becomes stronger under high current operation, the current density difference between the plain and the curved endplate becomes significant at high current density. As the PEMFC has larger active area, the current distribution is supposed to have poor uniformity owing to the pressure inhomogeneity. To compare the uniformity of the current distributions, the standard deviations of the measured current values of each current distribution are compared. As shown in Fig. 6(b), every standard deviation of the curved endplate is smaller than that of the plain endplate. The graph indicates that the PEMFC with the curved endplate is more uniformly activated than that with the plain endplate. For a detailed comparison between the clamping pressure and the contact resistance, the partial summation of the current and load were analyzed in the lateral direction. As shown in Fig. 7(a), the current distribution shows similar values in the lateral direction. Therefore, the regional current and exerted load were compared in the lateral direction. As the segmented cells of the PCB had a 7 16 composition, the establishments of these seven regions were reasonable. A current distribution of 700 mA/cm2, which was the highest current density value owing to the limitation
Fig. 5 e Current distributions of plain and curved endplates.
11682
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Fig. 6 e Current output comparison between PEMFCs with plain and curved endplates. (a) IeV curve of plain and curved endplates. (b) Standard deviation comparison of plain and curved endplates.
of the current sensor (AMETES CS2.5A-02, SENIS, Switzerland: maximum current density: 700 mA/cm2), was used for comparison. As shown in Fig. 7(b), the clamping load shows a “V”-shape, which is similar with current distribution. On the other hand, in Fig. 7(c), the clamping load shows a different distribution with the curved endplate in regions 1 and 7, indicating that the edge locations have lower load value than the center region. The result implies that a vertex depth of 0.5 mm is too high for achieving an equal load distribution. A more even load distribution can be obtained by controlling the vertex depth. Nevertheless, the standardization of the current values is significant for the even usage of the active area. Fig. 7(d) shows the average current of the segmented cell and the standard deviation under a current distribution of 700 mA/cm2. As the fuel cell system was operated under a constant current mode, the average current output is similar. However, the standard deviation of the current in the curved endplate shows a much lower value of 153 mA and that in the plain endplate shows a
Fig. 7 e Regional comparison between PEMFCs with plain and curved endplates. (a) Areal division for regional comparison. (b) Lateral sum of current and load in plain endplate. (c) Lateral sum of current and load in curved endplate. (d) Average and standard deviation of current distribution. value of 172 mA, which is 11% higher. Furthermore, the current sum of region 4 in the curved endplate is 34.1 A, whereas that in the plain endplate is 33.6 A. The power output and current evenness of the large-scale fuel cell system was
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
11683
Fig. 8 e Current distributions in parallel and counter flow modes.
improved without additional power consumption upon changing the shape of the endplate.
Gas feeding effect on current distribution To compare the effect of the gas feeding mode, the current distributions were obtained under two different feeding modes. Both experiments were performed using an identical single-cell PEMFC with a curved endplate. As shown in Fig. 8, the counter gas feeding mode shows a much narrower contour area in the left-center region of the active area. As the cell voltages are similar at the same current density, the counter flow mode shows higher activation uniformity than the parallel injecting mode. For quantitative analysis, the current measuring area is divided into areal proportions of 25%, 50%, and 25%, as shown in Fig. 9(a). If the local current densities are perfectly uniform, the total current in regions 1, 2, and 3 is directly proportional to the size of the active area. However, as shown in Fig. 9(b), the total current in each region shows a nonuniform distribution. Generally, the hydrogen inlet region (region 3) has ~1.1% and ~0.39% higher current proportion than the other side (region 1) under the parallel flow and the counter flow mode, respectively. Region 3 inevitably
has a larger current output than region 1 because the current density is high in the region adjacent to the hydrogen inlet. The standard deviations of the current densities are relatively low in the counter flow mode than in the parallel flow mode, as shown in Fig. 9(c). Although the overall current is the same, the higher uniformity of the current density is expected to ensure the improved durability of the PEMFC system.
Conclusions In this study, the current distributions were measured using a PCB-type segmented cell. A plain endplate with large active area showed an inhomogeneous pressure distribution. In this experiment, therefore, a curved endplate was used to increase the uniformity of the pressure distribution. The curved endplate increased the total pressed area, average pressure, and total load on a single-cell PEMFC. Furthermore, it improved the uniformity of the current distribution and total current density. At the same time, the effect of the gas feeding mode was also investigated. The experiments showed that the counter flow mode has higher current uniformity than the parallel flow mode. As the severe nonuniformity of the current
11684
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Fig. 9 e Current output comparison between parallel and counter flow modes. (a) Areal division for regional comparison. (b) Regional comparison between parallel and counter flows. (c) Standard deviation of current density in parallel and counter flows.
distribution is attributed to local degradation, a uniform current distribution is significant for the PEMFC performance and durability. Consequently, a more uniform pressure and gas distribution in the flow field are essential for better performance and durability. In this study, a passive method was
proposed for achieving a uniform pressure distribution without additional power consumption. The methodology of the present study can be used to measure the real-time status of the PEMFC by detecting the current density under various operation conditions.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Acknowledgements This work was supported by the New & Renewable Energy Core Technology Program (No. 2012T100100069) of the Korea Institute of Energy Technology Evaluation and Planning (KETEP), Republic of Korea.
references
[1] Steele BCH, Heinzel A. Materials for fuel-cell technologies. Nature 2001;410:345e52. [2] Zhang L, Liu Y, Song H, Wang S, Zhou Y, Hu SJ. Estimation of contact resistance in proton exchange membrane fuel cells. J Power Sources 2006;162:1165e71. [3] Ihonen J, Jaouen F, Lindbergh G, Sundholm G. A novel polymer electrolyte fuel cell for laboratory investigations and in-situ contact resistance measurements. Electrochim Acta 2001;146:2899e911. [4] Zhou P, Wu CW, Ma GJ. Contact resistance prediction and structure optimization of bipolar plates. J Power Sources 2006;159:1115e22. [5] Wen CY, Lin YS, Lu CH. Experimental study of clamping effects on the performance of a single proton exchange membrane fuel cell and a 10-cell stack. J Power Sources 2009;192:475e85.
11685
[6] Yu HN, Kim SS, Suh JD, Lee DG. Composite endplates with pre-curvature for PEMFC. Compos Struct 2010;92:1498e503. [7] Cleghorn SJC, Derouin CR, Wilson MS, Gottesfeld S. A printed circuit board approach to measuring current distribution in a fuel cell. J Appl Electrochem 1998;28:663e72. [8] Tang W, Lin R, Weng Y, Zhang J, Ma J. The effects of operating temperature on current density distribution and impedance spectroscopy by segmented fuel cell. Int J Hydrogen Energy 2013;38:10985e91. rez LC, Branda ~ o L, Sousa JM, Mendes A. Segmented [9] Pe polymer electrolyte membrane fuel cells-a review. Renew Sust Energ Rev 2011;15:169e85. [10] Wieser C, Helmbold A, Gu¨lzow E. A new technique for twodimensional current distribution measurement in electrochemical cells. J Appl Electrochem 2000;30(7):803e7. [11] Manso AP, Marzo FF, Barranco J, Garikano X, Mujika MG. Influence of geometric parameters of the flow fields on the performance of a PEM fuel cell. A review. Int J Hydrogen Energy 2012;37:15256e87. [12] Ge SH, Yi BL. A mathematical model for PEMFC in different flow modes. J Power Sources 2003;124:1e11. € ussler F, Zhang W, Ku¨ppers L, Jo € rissen L, [13] Scholta J, Ha Lehnert W. Development of a stack having an optimized flow field structure with low cross transport effects. J Power Sources 2006;155:60e5. [14] Sierra JM, Moreira J, Sebastian PJ. Numerical analysis of the effect of different gas feeding modes in a proton exchange membrane fuel cell with serpentine flow-field. J Power Sources 2011;196:5070e6.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Segmented cell approach for studying uniformity of current distribution in polymer electrolyte fuel cell operation Seung-Gon Kim, Min-Jin Kim, Young-Jun Sohn* Fuel Cell Laboratory, Korea Institute of Energy Research, 152, Gajeong-ro, Yuseong-gu, Daejeon 305-343, South Korea
article info
abstract
Article history:
In this study, the current distribution of a polymer electrolyte fuel cell (PEMFC) with a large
Received 13 February 2015
active area is investigated using a segmented cell system. A specially designed printed
Received in revised form
circuit board (PCB)-type segmented cell is applied to a single-cell PEMFC. By using the
6 May 2015
segmented system, the effects of clamping pressure uniformity between the components
Accepted 8 May 2015
of the PEMFC and the fuel injection direction are examined. The pressure uniformities of
Available online 10 June 2015
the two different types of endplates are measured using pressure indicating films (PIFs). A curved endplate is used to improve the pressure uniformity. The pressure uniformity is
Keywords:
found to significantly influence the current distribution in a PEMFC with a large active area.
Polymer electrolyte fuel cell
Two types of gas feeding modes, parallel and cross injection, are tested. The results show
Pressure uniformity
that the clamping pressure uniformity and gas feeding configuration affect the current
Segmented cell
distribution and overall PEMFC performance.
Pressure indicating film
Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Current distribution
Introduction Alternative power solutions such as wind power, solar power, and fuel cells have emerged owing to the growing demand for clean and renewable energy. Among these, polymer electrolyte membrane fuel cell (PEMFC) technology has attracted considerable attention as a mobile power solution owing to its relatively low operating temperature and suitable power range for vehicles [1]. However, the performance and durability of PEMFCs must be improved further before they can be commercialized. Generally, PEMFCs mainly suffer power loss owing to the contact resistance between the bipolar plates and the gas diffusion layers (GDLs) [2].
The contact resistance depends on the clamping pressure, gas pressure, current density, and temperature [3]. In Refs. [3], Ihonen et al. simultaneously measured the clamping pressure and contact resistance using a specially designed single cell. They validated the relationship between the contact resistance and other parameters through a comparison of in-situ and ex-situ measurements. Zhou et al. investigated the effect of the clamping force on the interfacial contact resistance and porosity of a GDL [4]. They suggested an optimal rib design for a gas channel for a reasonable combination of low interfacial contact resistance and good GDL porosity. Recently, pressure indicating films (PIFs) have been adapted to measure the pressure distribution. A PIF indicates
* Corresponding author. Tel.: þ82 42 860 3087; fax: þ82 42 860 3104. E-mail address: [email protected] (Y.-J. Sohn). http://dx.doi.org/10.1016/j.ijhydene.2015.05.055 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
11677
Fig. 1 e Segmented PEMFC. (a) Configuration of segmented PEMFC. (b) Photograph of PCB board.
applied pressure differences through color density variations, allowing the pressure distribution to be measured using software. Wen et al. investigated the fuel cell performance for different clamping torques and bolting configurations using PIFs. They found that a larger mean contact pressure leads to higher maximum power and uniformity of the contact pressure distribution. Furthermore, the ohmic resistance and mass transport limit current showed highly linear correlations with the mean contact pressure [5]. Yu et al. used a PIF to measure the pressure distribution in a single-cell PEMFC [6]. Instead of metals, they employed composite materials as endplates to increase the uniformity of the pressure distribution. They used composite endplates with a pre curvature created by thermal fabrication. These endplates red uced the stack weight without compromising the performance. In this study, PIF was used to measure the pressure distribution in a single-cell PEMFC. A curved endplate was used
to achieve high pressure uniformity. The effect of the pressure distribution was analyzed based on the current distribution. We adapted a printed circuit board (PCB)-type segmented cell to visualize the current distribution. Segmented cells were generally used to measure the current distributions [7,8]. Three techniques were used to realize a segmented cell system: PCB, resistor network, and Hall effect sensors [9]. Cleghorn et al. [7] first adapted the PCB-based measuring technique, in which a segmented flow field and a current collector were used. Wieser et al. [10] first introduced the Hall effect sensor for indirect current sensing. They investigated the relation between the uniformity of the pressure and the current distribution using a PCB-type segmented cell. As mass transport is one of the most significant phenomena in PEMFCs owing to the electrochemical reaction, many studies have investigated efficient flow configurations through numerical as well as experimental methods [11]. However, only a
11678
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
few studies have investigated the gas feeding modes. Ge and Yi investigated the influence of the gas feeding mode in a PEMFC fuel cell with an active area of 140 cm2 [12]. They compared the counter flow and the co-flow in terms of the membrane humidification and current density distribution. Scholta et al. studied the PEMFC performance for different gas feeding modes
of the co-flow, counter flow, and cross flow [13]. Recently, Sierra et al. classified the gas feeding configuration into four modes: cross flow, nonsymmetrical flow, similar flow, and counter flow [14]. They reported that the local current distribution in the catalyst layer can be improved by changing the gas feeding mode. In this study, the current distributions with different gas
Fig. 2 e Experimental setup of segmented PEMFC. (a) Schematic of experimental setup. (b) Photograph of experimental setup.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
feeding modes are obtained using a segmented cell system. As a uniform current distribution leads to a uniform temperature and liquid water production [11], better performance and durability can be achieved with a uniform pressure distribution and optimized gas feeding configuration.
Experimental method and apparatus Experimental setup Fig. 1(a) shows the configuration of a single-cell PEMFC with a PCB-type segmented cell. A segmented PCB with 112 (7 16) segments was installed between a gas flow channel and an endplate. The size of each segment was 1.79 cm 1.79 cm. Metal-build gas flow channels with a large active area of 360 cm2 were employed. The PCB collects the electric current generated by the PEMFC. Fig. 1(b) shows a photograph of the segmented PCB used in this experiment. Fig. 2(a) shows a schematic diagram of the experimental setup. The local current from the PEMFC passes through each segmented cell of the PCB. The current is distributed to the current sensors (AMETES CS2.5A-02, SENIS, Switzerland).
11679
These sensors are connected in the unipolar operation mode (0e5 V), corresponding to a current of 0e2.5 A (0e780 mA/ cm2) for each sensor. Therefore, the maximum current density was limited to 700 mA/cm2 for ensuring the safety of the system. The distributed current was collected to the electric load (PLZ664WA, KIKUSUI, Japan). Owing to the electrochemical characteristics of a single-cell PEMFC with a large active area, the electric load needs to be operated under low voltage (below 1 V) and high current (of the order of hundreds of amperes). Fig. 2(b) shows a photograph of the experimental setup.
Segmented cell calibration Although all segments are installed in the PCB, each electric path from the segmented cell surface to the data acquisition (DAQ) system varies. To compensate the ohmic difference between the PCB and the DAQ systems, a calibration procedure was performed using a calibration unit, as shown in Fig. 3(a). The calibration process was performed as follows. Fig. 3(b) shows a flowchart of the calibration process.
Fig. 3 e Current calibration method. (a) Segmented PCB calibration unit. (b) Flowchart of segmented PCB calibration. (c) Current calibration using four-point least squares method.
11680
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Fig. 4 e Pressure distribution of plain and curved endplates.
1. Charge the reference currents to the first segmented cell 2. From the measured values, formulate a transposing equation using linear squares fit. 3. Move to the next segmented cell. 4. Repeat steps 1e3 for every segmented cell until the last segmented cell. 5. Apply every transposing equation during current acquisition in the experiment. In this experiment, by applying the accurate current induced by the DC power supply as a reference value, the measured value can be corrected to the calibrated value using a least squares method. Four different current values (0, 200, 400, and 600 mA) were charged to a single segmented cell as reference currents. Fig. 3(c) shows the typical calibration result obtained using the least squares method. A solid line shows the current actually charged in a segmented cell by the DC power supply. However, the value measured by the current sensor and data acquisition system differ from the actual value. These gaps between the actual and the measured values increase at higher current density. After the calibration process described above, the calibrated values approach the actual current values.
Results and discussion Pressure distribution Two types of endplates were employed to compare the pressure distributions, which were of the plain and the curved type. Both endplates were made of 30-mm-thick epoxy glass. Fourteen M10 bolts were used for the combination of endplates. All bolts were fastened with the same torque of 110 kgf$cm. In the case of a curved endplate, the vertex depth was 0.5 mm. Fig. 4 shows the pressure distribution using a PIF. A
Table 1 e Pressure distributions according to the shape of endplate. End-plates shape
Plain end-plate
Curved end-plate
Pressed Area [mm2] Ave. Pressure [Mpa] Max. Pressure [Mpa] Load [N]
27761 0.77 3.06 21321
36416 0.96 3.06 34970
22 C, 46%RH, Continuous pressing mode.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
“Super Low” PIF (FUJIFILM, Japan) was used during the experiment at pressures of 0.5e2.5 MPa [11]. Quantitative pressure information was obtained using an FPD-8010E pressure distribution mapping system (FUJIFILM, Japan). Table 1 summarizes the experimental data. As shown in Table 1, the total pressed area and average pressure of the curved endplate is much larger than that of the plain endplate, although the maximum pressures have the same value. This result indicates that the curved endplate efficiently pressed the single-cell PEMFC.
Curvature effect of endplate on current distribution Fig. 5 shows the typical current distribution results of a singlecell PEMFC. The center region of the curved endplate shows higher activation status than the plain endplate. Owing to these activation differences, the cell voltage of the curved endplate is higher at the same current density than the plain endplate. Fig. 6(a) shows the overall current density curve of the two different types of endplates. As shown in the graph, the current density of the plain endplate rapidly decreased in the high current density region. This was mainly attributed to
11681
increased ohmic contact resistance owing to the pressure distribution. As the effect of the ohmic resistance becomes stronger under high current operation, the current density difference between the plain and the curved endplate becomes significant at high current density. As the PEMFC has larger active area, the current distribution is supposed to have poor uniformity owing to the pressure inhomogeneity. To compare the uniformity of the current distributions, the standard deviations of the measured current values of each current distribution are compared. As shown in Fig. 6(b), every standard deviation of the curved endplate is smaller than that of the plain endplate. The graph indicates that the PEMFC with the curved endplate is more uniformly activated than that with the plain endplate. For a detailed comparison between the clamping pressure and the contact resistance, the partial summation of the current and load were analyzed in the lateral direction. As shown in Fig. 7(a), the current distribution shows similar values in the lateral direction. Therefore, the regional current and exerted load were compared in the lateral direction. As the segmented cells of the PCB had a 7 16 composition, the establishments of these seven regions were reasonable. A current distribution of 700 mA/cm2, which was the highest current density value owing to the limitation
Fig. 5 e Current distributions of plain and curved endplates.
11682
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Fig. 6 e Current output comparison between PEMFCs with plain and curved endplates. (a) IeV curve of plain and curved endplates. (b) Standard deviation comparison of plain and curved endplates.
of the current sensor (AMETES CS2.5A-02, SENIS, Switzerland: maximum current density: 700 mA/cm2), was used for comparison. As shown in Fig. 7(b), the clamping load shows a “V”-shape, which is similar with current distribution. On the other hand, in Fig. 7(c), the clamping load shows a different distribution with the curved endplate in regions 1 and 7, indicating that the edge locations have lower load value than the center region. The result implies that a vertex depth of 0.5 mm is too high for achieving an equal load distribution. A more even load distribution can be obtained by controlling the vertex depth. Nevertheless, the standardization of the current values is significant for the even usage of the active area. Fig. 7(d) shows the average current of the segmented cell and the standard deviation under a current distribution of 700 mA/cm2. As the fuel cell system was operated under a constant current mode, the average current output is similar. However, the standard deviation of the current in the curved endplate shows a much lower value of 153 mA and that in the plain endplate shows a
Fig. 7 e Regional comparison between PEMFCs with plain and curved endplates. (a) Areal division for regional comparison. (b) Lateral sum of current and load in plain endplate. (c) Lateral sum of current and load in curved endplate. (d) Average and standard deviation of current distribution. value of 172 mA, which is 11% higher. Furthermore, the current sum of region 4 in the curved endplate is 34.1 A, whereas that in the plain endplate is 33.6 A. The power output and current evenness of the large-scale fuel cell system was
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
11683
Fig. 8 e Current distributions in parallel and counter flow modes.
improved without additional power consumption upon changing the shape of the endplate.
Gas feeding effect on current distribution To compare the effect of the gas feeding mode, the current distributions were obtained under two different feeding modes. Both experiments were performed using an identical single-cell PEMFC with a curved endplate. As shown in Fig. 8, the counter gas feeding mode shows a much narrower contour area in the left-center region of the active area. As the cell voltages are similar at the same current density, the counter flow mode shows higher activation uniformity than the parallel injecting mode. For quantitative analysis, the current measuring area is divided into areal proportions of 25%, 50%, and 25%, as shown in Fig. 9(a). If the local current densities are perfectly uniform, the total current in regions 1, 2, and 3 is directly proportional to the size of the active area. However, as shown in Fig. 9(b), the total current in each region shows a nonuniform distribution. Generally, the hydrogen inlet region (region 3) has ~1.1% and ~0.39% higher current proportion than the other side (region 1) under the parallel flow and the counter flow mode, respectively. Region 3 inevitably
has a larger current output than region 1 because the current density is high in the region adjacent to the hydrogen inlet. The standard deviations of the current densities are relatively low in the counter flow mode than in the parallel flow mode, as shown in Fig. 9(c). Although the overall current is the same, the higher uniformity of the current density is expected to ensure the improved durability of the PEMFC system.
Conclusions In this study, the current distributions were measured using a PCB-type segmented cell. A plain endplate with large active area showed an inhomogeneous pressure distribution. In this experiment, therefore, a curved endplate was used to increase the uniformity of the pressure distribution. The curved endplate increased the total pressed area, average pressure, and total load on a single-cell PEMFC. Furthermore, it improved the uniformity of the current distribution and total current density. At the same time, the effect of the gas feeding mode was also investigated. The experiments showed that the counter flow mode has higher current uniformity than the parallel flow mode. As the severe nonuniformity of the current
11684
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Fig. 9 e Current output comparison between parallel and counter flow modes. (a) Areal division for regional comparison. (b) Regional comparison between parallel and counter flows. (c) Standard deviation of current density in parallel and counter flows.
distribution is attributed to local degradation, a uniform current distribution is significant for the PEMFC performance and durability. Consequently, a more uniform pressure and gas distribution in the flow field are essential for better performance and durability. In this study, a passive method was
proposed for achieving a uniform pressure distribution without additional power consumption. The methodology of the present study can be used to measure the real-time status of the PEMFC by detecting the current density under various operation conditions.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 1 6 7 6 e1 1 6 8 5
Acknowledgements This work was supported by the New & Renewable Energy Core Technology Program (No. 2012T100100069) of the Korea Institute of Energy Technology Evaluation and Planning (KETEP), Republic of Korea.
references
[1] Steele BCH, Heinzel A. Materials for fuel-cell technologies. Nature 2001;410:345e52. [2] Zhang L, Liu Y, Song H, Wang S, Zhou Y, Hu SJ. Estimation of contact resistance in proton exchange membrane fuel cells. J Power Sources 2006;162:1165e71. [3] Ihonen J, Jaouen F, Lindbergh G, Sundholm G. A novel polymer electrolyte fuel cell for laboratory investigations and in-situ contact resistance measurements. Electrochim Acta 2001;146:2899e911. [4] Zhou P, Wu CW, Ma GJ. Contact resistance prediction and structure optimization of bipolar plates. J Power Sources 2006;159:1115e22. [5] Wen CY, Lin YS, Lu CH. Experimental study of clamping effects on the performance of a single proton exchange membrane fuel cell and a 10-cell stack. J Power Sources 2009;192:475e85.
11685
[6] Yu HN, Kim SS, Suh JD, Lee DG. Composite endplates with pre-curvature for PEMFC. Compos Struct 2010;92:1498e503. [7] Cleghorn SJC, Derouin CR, Wilson MS, Gottesfeld S. A printed circuit board approach to measuring current distribution in a fuel cell. J Appl Electrochem 1998;28:663e72. [8] Tang W, Lin R, Weng Y, Zhang J, Ma J. The effects of operating temperature on current density distribution and impedance spectroscopy by segmented fuel cell. Int J Hydrogen Energy 2013;38:10985e91. rez LC, Branda ~ o L, Sousa JM, Mendes A. Segmented [9] Pe polymer electrolyte membrane fuel cells-a review. Renew Sust Energ Rev 2011;15:169e85. [10] Wieser C, Helmbold A, Gu¨lzow E. A new technique for twodimensional current distribution measurement in electrochemical cells. J Appl Electrochem 2000;30(7):803e7. [11] Manso AP, Marzo FF, Barranco J, Garikano X, Mujika MG. Influence of geometric parameters of the flow fields on the performance of a PEM fuel cell. A review. Int J Hydrogen Energy 2012;37:15256e87. [12] Ge SH, Yi BL. A mathematical model for PEMFC in different flow modes. J Power Sources 2003;124:1e11. € ussler F, Zhang W, Ku¨ppers L, Jo € rissen L, [13] Scholta J, Ha Lehnert W. Development of a stack having an optimized flow field structure with low cross transport effects. J Power Sources 2006;155:60e5. [14] Sierra JM, Moreira J, Sebastian PJ. Numerical analysis of the effect of different gas feeding modes in a proton exchange membrane fuel cell with serpentine flow-field. J Power Sources 2011;196:5070e6.