Elsevier

Journal of Power Sources

Volume 412, 1 February 2019, Pages 717-724
Journal of Power Sources

In-situ measurement of temperature and humidity distribution in gas channels for commercial-size proton exchange membrane fuel cells

https://doi.org/10.1016/j.jpowsour.2018.12.008Get rights and content

Highlights

  • Measurement with micro-sensors for the commercial-size fuel cell is effective.

  • Uneven temperature and humidity distribution in the cathode flow field is obtained.

  • Performance and resistance are greatly related to temperature-humidity distribution.

  • Counter-flow without humidification gets good performance and water-heat balance.

Abstract

It is necessary to obtain internal heat and water distribution of proton exchange membrane fuel cells in order to perform water-heat management. Most of the existing works focus on the measurement in small experimental fuel cells, which cannot effectively guide the development of large-area fuel cells. In this study, an in-situ measurement method using micro-sensors is developed to observe the temperature and relative humidity in a commercial-size fuel cell with parallel flow fields and an active area of 250 cm2. A sensor array is incorporated into the cathode flow field, and well-designed waterproof and sealing structures protect sensors from liquid water. The sensors are verified with accuracy and response speed. The fuel cell performance and electrochemical impedance spectroscopy under co-flow and counter-flow configurations considering anode gas humidification are investigated. The results show that the temperature distribution is more uneven in the cases with lower output voltage, and the fuel cell performance is significantly affected by the humidity near the air inlet area. As current density increases, the relative humidity drops while dew point temperature keeps almost unchanged. The experimental method and analysis are beneficial to the understanding of the water-heat state and optimization of fuel cell stack design.

Introduction

Proton exchange membrane fuel cells (PEMFCs) show bright prospect in automobiles, unmanned aerial vehicles and stationary power systems of the future owing to their advantages such as low pollution, low noise, low operating temperature, high power density and high energy conversion efficiency [1,2].

Water-heat management is a critical issue in the operation of PEMFCs and affects the performance and durability [[3], [4], [5], [6]]. Water is the product of the electrochemical reaction, and it also comes from humidified gas. Inside the fuel cell, water transfers from the anode to the cathode due to the electro-osmotic drag, and water concentration diffusion occurs. Some water remains in the membrane electrode assemblies (MEAs) and the flow field to maintain the water balance, and the excess water is discharged through the gas channels. An appropriate quantity of water contributes to the high proton conductivity of the membrane, and excess water increases the resistance of mass transfer. The saturated vapour pressure of water varies at different temperatures, so the temperature determines the content of liquid water in the fuel cell. Hence, water, heat and the electrochemical reaction need a suitable balance to reach high performance in the fuel cell. It is necessary to obtain the balance status of the fuel cell in order to perform fuel cell optimization.

The state of water content and temperature during the electrochemical process can be obtained by means of external or internal measurements of the fuel cell. External measurement methods include installing sensors on the gas pipelines, analysis of discharged water and thermal imaging of the outer surface of the fuel cell. Due to the limited information obtained from external measurements, researchers have taken a variety of experimental methods to measure the internal water-heat state of fuel cells. Visualization with high-speed photography is most commonly used to observe the two-phase transport in channels [7]. Imaging methods including X-ray [8,9], MRI [10] and neutron imaging [11,12] are used to detect the water thickness in the fuel cell. Besides, varieties of sensors are developed to directly measure the temperature and water content inside the fuel cells. Fibre Bragg grating (FBG) sensors [13,14], tunable diode laser absorption spectroscopy (TDLAS) [15], thin film temperature sensors [16], bandgap temperature sensors and capacitative humidity sensors [[17], [18], [19], [20]] have been shown with a repeatable and reliable response.

These works help to understand the distribution of temperature and water, but they cannot thoroughly guide an actual commercial fuel cell design and development. Most of the existing works developed small experimental fuel cells with an active area ranging from 5 to 50 cm2. However, the commercial fuel cells usually have an active area of 200–400 cm2, leading to differences in heat dissipation, reaction gas distribution and water transfer characteristic. In addition, the flow field types considered in the existing works are limited to serpentine and interdigitated configurations. Instead, the parallel flow field is usually used in the commercial fuel cells due to its low pressure drop for reactants, especially for the air. The parallel channels are more conducive to the uniform distribution of the reactants and the prevention of local water plugging [[21], [22], [23]]. Moreover, the above measurement methods have their limitations. For example, they are impractical to measure at a specific local point and to arrange densely in the flow field. Water droplets in gas channels may block the optical path or sensing window of these sensors.

In this study, the authors aim at in-situ measuring the temperature and humidity distributions in a commercial-size large-scale fuel cell for vehicle application. The fuel cell has an active area of 250 cm2 and a flow field with 40 parallel channels. The purpose is to study the temperature-humidity balance at the gas channel level and to understand the inhomogeneity between these parallel channels. The authors expect to measure the temperature and humidity at the same time to obtain a large area distribution and minimize the additional impact of measurement. A novel experimental fuel cell design is proposed with numbers of commercial temperature and humidity micro-sensors. Disadvantages existed in works of Hinds et al. [24,25] are solved by considering isolation of liquid water and arrangement of sensors inside the flow field. The experimental result reveals that the temperature and humidity in the flow field are uneven and significantly affected by gas humidification, flow direction, and current density.

Section snippets

Experimental fuel cell

The experimental fuel cell is well-designed and assembled for the measurement of temperature and relative humidity distribution inside the cathode flow field. The experimental fuel cell takes into consideration the arrangement of the sensors, and relevant structural changes are made on the basis of a conventional fuel cell without sensors and sensor installation structures.

Fig. 1 (a) shows the schematic diagram of the cathode flow field plate made of graphite. One side of the plate is the gas

Temperature and relative humidity distribution

In order to visualize the measurement, temperature and humidity values are arranged according to the relative positions of the measuring points and drawn into a colour matrix image. The temperature and relative humidity distribution of case #2 at a current density of 1.0 A cm−2 is taken as the example as shown in Fig. 4. Values in column 0 and column 6 stands for the measurement points in the diversion area and the confluence area. For each row, column 1–5 are on the same channel. Air flows

Comparison of fuel cell performance and resistance

The polarization curves measured in case #1-#4 are plotted in Fig. 5 (a). The result shows that case #1-Counter-humidified obtains the maximum performance and case #4-Co-dry gets the lowest power output. The polarization curves are divided into three segments according to the current density. In low current range (<0.4 A cm−2), hydrogen and air are supplied at high flow rates corresponding to the current density of 0.4 A cm−2, and generated water is not enough to maintain the membrane

Conclusion

In the present study, a commercial-size large-scale experimental fuel cell with parallel flow fields and an active area of 250 cm2 is well-designed and assembled. Sensirion SHT series temperature and humidity sensors are incorporated into the graphite flow field plate to measure the temperature and relative humidity distribution in the cathode flow field. The conclusion can be drawn as follows.

  • 1)

    Temperature and relative humidity distribution in the flow field are uneven. The outside temperatures

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 51705308) and National Key Research and Development Program of China (2017YFB0102900).

References (29)

  • T. Kotaka et al.

    Electrochim. Acta

    (2014)
  • X. Liu et al.

    Electrochim. Acta

    (2007)
  • F. Bi et al.

    Int. J. Hydrogen Energy

    (2015)
  • P. Yi et al.

    J. Power Sources

    (2013)
  • I.S. Hussaini et al.

    J. Power Sources

    (2009)
  • J. Lee et al.

    J. Power Sources

    (2013)
  • T. Arlt et al.

    J. Power Sources

    (2014)
  • R. Satija et al.

    J. Power Sources

    (2004)
  • M. Weiland et al.

    Electrochim. Acta

    (2013)
  • N. David et al.

    Int. J. Hydrogen Energy

    (2014)
  • S. He et al.

    Sensor Actuator Phys.

    (2006)
  • C.-Y. Lee et al.

    J. Power Sources

    (2007)
  • A. Su et al.

    Int. J. Hydrogen Energy

    (2012)
  • Y. Xu et al.

    Int. J. Hydrogen Energy

    (2016)
  • Cited by (57)

    View all citing articles on Scopus
    View full text