Fuel Cell Water Balance Research Articles

Water Balance Measurements for a Single Proton Exchange Membrane Fuel Cell Using a Hot Wire Anemometer

Water balance data for a single proton exchange membrane fuel cell with an active area of 50 cm2 were measured with a hot wire anemometer, placed in the anode outlet stream. The measured hot wire voltage, E, is divided by the calibration signal, E_0, that was taken for a dry hydrogen stream of the same molar flow as the anode outlet at the same temperature. Because water vapor has a much lower thermal conductivity than hydrogen, the recorded hot wire voltage is always lower than the calibration signal, and there is a clear correlation between the relative voltage signal E/E0 and the fuel cell water balance. The relative voltage single can be easily computed at any point in time to thus provide a real-time signal of the fuel cell water balance.This data was for a variety of operating conditions. The results show the same trends that were observed by other researchers who used the water knock-out method, only with higher accuracy. A decrease in the anode inlet relative humidity decreases the water balance value, rd , significantly while the cathode inlet relative humidity only slightly affects the measured water balance. The membrane-electrode-assembly in this work was provided by Hydrogen South Africa – Catalysis (HySA Catalysis). Figure 1

Water Balance Measurements for a Single Proton Exchange Membrane Fuel Cell Using a Hot Wire Anemometer

Recently, our group has developed a novel technique to measure the water balance in a proton exchange membrane fuel cell with a hot wire anemometer. This device yields a high-frequency voltage signal that can be converted into the fuel cell water balance, and thus producing almost real time water balance data that can complete the usual experimental data set consisting of cell voltage, current, and high frequency resistance. In this work, we are presenting water balance data of a single 50 cm2 fuel cell with pre-humidified anode and cathode streams. The results show a clear tendency of anode dehydration at elevated current densities which appears to be limiting the current density that can be drawn from the cell.

A Comparison of the Proton Exchange Membrane Fuel Cell Water Balance Measured Simultaneously with the Water Knock-out Method and Hot Wire Anemometry

The water balance of a proton exchange membrane fuel cell (PEMFC) can provide valuable insight into the state of the fuel cell and may even limit the current density due to anode dry-out. Typically, commercial fuel cell test stations can be equipped with a method of detecting the fuel cell water balance using the water knock-out method. In this method, the anode and/or cathode exhaust gas streams are led through a condensing unit and the liquid water that is collected over time is back-calculated in terms of the water balance, usually denoted rd and defined below. Thus, it is a dimensionless property that indicates how much product water has been transported through the membrane from anode to cathode, and negative values indicate overall back-transport from cathode to anode. The definition is also such that when multiplying with the factor 2 one obtains the percentage of the product water that has crossed the membrane, i.e. a value of -0.02 indicates that 4% of the product water has been back-transported from cathode to anode. Adding the option to measure the water balance to fuel cell test stations increases the cost significantly because additional equipment has to be added to the system such as heat exchangers, through-liquid drainers, beakers, mass scales, control valves and all the hardware and software for controlling and integrating the water knock-out system. After the gas stream escapes from the cathode and anode outlet it passes through the heat exchanger in order to cool the gas stream and condense the water in the stream. The liquid drainer delivers the product water to a beaker on a mass balance scale. The software monitors the mass until the water level reaches a high point, and then drains the water automatically through a solenoid (GREENLIGHT INNOVATION, Canada). Because there is no external cooling reservoir, the condensing of water occurs at room temperature and some water vapor escapes the measurement. In addition, such an experiment has to run for an extended period of time in order to collect a sufficient amount of water for an accurate result. Meanwhile, our research group has developed a novel method to measure the fuel cell water balance directly and in real time using hot wire anemometry. This method yields a voltage signal that can be directly converted into the fuel cell water balance such that the fuel cell water balance can be a direct output signal similar to voltage and current density. Because, this method requires the hot wire and a heating coil to control the temperature as an extra added hardware in the anode outlet. In addition, it provides an ad-hoc real time electrical signal of the PEMFC water balance, thus the cell size and the current density have no significant effect on the accuracy of measured water balance. The Ad-hoc real time rd measurements allow this method to be part of FCV systems and/or their test equipment. Behind the test section the anode exhaust gas is just led through the fume hood and o drainage equipment is required.In this work, the water balance for 5 KW water-cooled PEMFC stack is simultaneously measured using the two aforementioned methods. The stack is run under anode and cathode wet conditions at different current densities and stoichiometric ratios. The hot-wire method has shown a higher sensitivity and an instant response for the measured rd when changing the running conditions such as the current density. Meanwhile, the knock-out method showed less sensitivity and much slower time response for the measured rd. In addition, it suffers from higher inaccuracies due the loss water vapor with the exhausted gases. However, Figure 1 shows that the overall comparison between both methods is very good and may serve as validation of the hot wire method. Figure 1

A Comparison of the Proton Exchange Membrane Fuel Cell Water Balance Measured Simultaneously with the Water Knock-out Method and Hot Wire Anemometry

The water balance of a proton exchange membrane fuel cell (PEMFC) can provide valuable insight into the state of the fuel cell. The most commonly applied method to measure the water balance is the knock-out method where the flaw gas of anode and/or cathode is cooled down and the condensed amount of water is weighed and compared to the amount of product water, while over the past years our group has developed a new method by applying hot wire anemometry at the anode outlet. In this work, a comparison between the knock-out and hot wire methods is experimentally demonstrated for 3.8 kW water-cooled PEMFC stack. The stack is run under wet conditions at different current densities and stoichiometric ratios. The hot-wire method has shown a higher sensitivity and an instant response for the measured water balance. Meanwhile, the knock-out method showed less sensitivity, much slower time response, and higher inaccuracies.

Proton exchange membrane fuel cell diagnosis by spectral characterization of the electrochemical noise

Electrochemical noise analysis (ENA) has been performed for the diagnosis of proton-exchange membrane fuel cell (PEMFC) under various operating conditions. Its interest is related with the possibility of a non-invasive on-line diagnosis of a commercial fuel cell. A methodology of spectral analysis has been developed and an evaluation of the stationarity of the signal has been proposed. It has been revealed that the spectral signature of fuel cell, is a linear slope with a fractional power dependence 1/fα where α = 2 for different relative humidities and current densities. Experimental results reveal that the electrochemical noise is sensitive to the water management, especially under dry conditions. At RHH2 = 20% and RHair = 20%, spectral analysis shows a three linear slopes signature on the spectrum at low frequency range (f < 100 Hz). This results indicates that power spectral density, calculated thanks to FFT, can be used for the detection of an incorrect fuel cell water balance.

The Effect of Nitrogen Cross-Over on Water Balance Measurements in Proton Exchange Membrane Fuel Cell Using Constant Temperature Anemometry

R&D on Proton exchange membrane fuel cells (PEMFC’s) technology has been accelerated in the last few years to reduce the dependency on fossil fuel. PEMFC’s operate by internally combining oxygen from air with hydrogen to form water and generate electricity and waste heat. The most common hurdle for enhanced PEMFC durability and performance is still the water management: the proton exchange membrane in the center of these fuel cells has to be hydrated in order to keep its ability to conduct protons from anode to cathode side while on the other hand excessive liquid water can lead to cell flooding and increased degradation rates of the cell. Thus, a detailed understanding of all aspects of water management in PEMFC is important. This includes the fuel cell water balance, i.e. the question which fraction of the product water leaves the fuel cell via the anode channels versus the cathode channel. Our research group is currently developing a state of art technology to obtain an ad-hoc and real time electrical signal of the fuel cell water balance by employing a constant temperature hot wire anemometry [1]. The hot wire sensor is placed in the anode outlet of PEMFC, and the voltage signal received gives valuable insight into heat and mass transfer phenomena which can be interpreted directly to real time PEMFC water balance. So far, ex-situ experiments have been conducted to measure the voltage signal of the hot wire anemometer for a known gas stream that contains a binary mixture of hydrogen and water vapour as would ideally leave the fuel cell anode. However, in an operating fuel cell there are two main uncertainties in this method: (i) there is some internal hydrogen crossover that does not load to an external current, thus the exact amount of hydrogen leaving the fuel cell anode is not known, and (ii) that there is nitrogen crossing over from the cathode side to the anode side, and this nitrogen does slighty falsify the voltage signal that the hot wire yields. The effects of nitrogen-cross over on the hot wire voltage signal at the anode outlet and consequently on the measured water balance (rd ) is studied in this work. In our previous work it was shown that the only unknown in the determination of the hot wire voltage signal is the equation that determines the heat transfer around the hot wire, and we have shown that it is necessary to employ a power-law equation as suggested by Hilpert: Nu = C Pr0.33 Rem where Re is the Reynolds number and Pr is the Prandtl number are calculated at film temperature. The constants C and m usually depend on the Re, and they will be determined out of the experimental data for the gas mixture, by plotting the measured (Nu/ Pr0.333 ) versus the Re number of the gas mixture stream. The missing C and m from Hilpert equation can then be easily obtained from a power-fit of the general form Y = CXm . Also, it was shown somewhere else that in fact only the exponent to the Reynolds number, of the gas mixture stream is important when determining the fuel cell water balance out of the hot wire signal [1]. The nitrogen-cross over is experimentally demonstrated by introducing 1% of nitrogen to the dry hydrogen molar flow. The 99%H2 +1%N2 of the dry mixture is humidified with water vapor by controlling the relative humidity (RH) of the dry mixture with in the range of (0-100)%RH, simulating the PEMFC anode outlet. The hot wire voltage is measured with and without nitrogen and it was slightly lower with the presence of 1 % nitrogen in the flow. The effect of the voltage reduction on the measured water balance can be neglected. This because the effect of 1% nitrogen on power law constant’s m which is used in determining the water balance as explained somewhere else is extremely low. It might be concluded that, the hot wire technique for measuring the water balance is still accurate and can be used for ad-hoc and real time water balance measurements and the nitrogen cross-over affect can be neglected, knowing that the 1% N2 is one order of magnitude higher than the actual concertation of 0.02 m2 active area cell used for lab testing. [1] Berning, T., & Al Shakhshir, S., Applying hot wire anemometry to directly measure the water balance in a proton exchange membrane fuel cell–Part 1: Theory. International Journal of Hydrogen Energy, 40(36) (2015), 12400-12412.

Applying hot-wire anemometry to directly measure the water balance in a proton exchange membrane fuel cell – Part 2: Experimental

In order to better understand and more accurately measure the water balance in a proton exchange membrane fuel cell our group has recently proposed to apply hot wire anemometry in the fuel cell's anode outlet. It was theoretically shown that the electrical signal obtained from the hot wire sensor can be directly converted into the fuel cell water balance. In this work an ex-situ experimental investigation is performed to examine the effect of the wire diameter and the outlet pipe diameter on the voltage signal. For a laboratory fuel cell where the mass flow rate the anode outlet is small, it is found important to use a small output pipe diameter to obtain a sufficiently strong convection effect and hence clear voltage readings. Depending on the hot wire diameter and the inner pipe diameter, the resulting values for the exponent of the Reynolds number Re in the determination of the Nusselt number Nu range between m = 0.267 and m = 0.329. In general, it is shown that applying hot wire anemometry yields in fact very clear voltage readings with high frequency, and it can be used as a diagnosis tool in various fuel cell applications.

Applying hot-wire anemometry to directly measure the water balance in a proton exchange membrane fuel cell for a pre-humidified hydrogen stream

In a recent publication it has been shown how the water balance in a proton exchange membrane fuel cell can be determined employing hot wire anemometry. The hot wire sensor has to be placed into the anode outlet pipe of the operating fuel cell, and the voltage signal E that is read from the sensor has to be divided by a pre-determined voltage signal E0 that has been obtained for a stream of dry hydrogen where the molar flow rate corresponds to a total current I of the fuel cell stack and a stoichiometric flow ratio, ξ. Because the last two properties are usually continuously known in fuel cell experiments, E0 is also continuously known. There is a one-to-one correlation between the relative voltage signal E/E0 and the fuel cell water balance, and therefore the fuel cell water balance can be a continuous output signal similar to the fuel cell voltage and the high frequency resistance. This method was originally believed to be limited to the fuel cell anode operating on dry hydrogen. In the current work, it is expanded for the case of a pre-humidified hydrogen stream. In addition, useful correlations are derived that link the fuel cell water balance with the anode side inlet and outlet thermodynamic state. Finally, it will be shown how previously developed dew point diagrams for the anode side in a fuel cell can be corrected for a humidified hydrogen inlet stream.

Modeling and experimental validation of water mass balance in a PEM fuel cell stack

Polymer electrolyte membrane (PEM) fuel cells require good hydration in order to deliver high performance and ensure long life operation. Water is essential for proton conductivity in the membrane which increases by nearly six orders of magnitude from dry to fully hydrated. Adequate water management in PEM fuel cell is crucial in order to avoid an imbalance between water production and water removal from the fuel cell.In the present study, a novel mathematical zero-dimensional model has been formulated for the water mass balance and hydration of a polymer electrolyte membrane. This model incorporates all the essential fundamental physical and electrochemical processes occurring in the membrane electrolyte and considers the water adsorption/desorption phenomena in the membrane. The effect of diffusivity model, surface roughness and water content driving force is considered. The model is validated against experimental data. In the results it is shown that the fuel cell water balance calculated by this model shows better fit with experimental data-points compared with model where only steady state operation were considered. We conclude that this discrepancy is due a different rate of water transport when membrane absorption/desorption is considered in the model. The model becomes useful in system modeling when studying fuel cells in dynamic conditions.

Applying hot wire anemometry to directly measure the water balance in a proton exchange membrane fuel cell – Part 1: Theory

In order to accurately determine the water balance of a proton exchange membrane fuel cell it has recently been suggested to employ constant temperature anemometry (CTA), a frequently used method to measure the velocity of a fluid stream. CTA relies on convective heat transfer around a heated wire at around 200 °C. The heat loss to the fluid stream is balanced by electrical power dissipation, and the required voltage E is the output signal which is highly sensitive to the heat transfer coefficient of the wire and therefore provides accurate readings. This work explains the theory and summarizes the equations required to calculate the heat transfer coefficient and the resulting voltage signal as function of the fuel cell water balance. The most critical and least understood part is the determination of the Nusselt number to calculate the heat transfer between the wire and the gas stream. Different expressions taken from the literature will be examined in detail, and it will be demonstrated that the power-law approach suggested by Hilpert is the only useful one for the current purposes because in this case the voltage response from the hot-wire sensor E/E0 shows the same dependency to the water balance for all current densities. Therefore, only one curve-fit equation will be required. The voltage curve E0 is an arbitrary calibration curve, and this can be conveniently chosen to be the voltage signal for a dry hydrogen stream at a given temperature and various flow rates which can be easily measured.

Modelling and Validation of Water Hydration of PEM Fuel Cell Membrane in Dynamic Operations

Fuel cells convert chemical energy of hydrogen and oxygen directly in to electricity. High efficiency and zero emission have made them a prime candidate for powering the next generation of electric vehicles. PEM fuel cells require good hydration in order to deliver high performance and ensure long life operation. Water is essential for membrane proton conductivity which increases by nearly six orders of magnitude from dry to fully hydrated. Adequate water management in PEM fuel cell is crucial in order to avoid an imbalance between water production and water removal from the fuel cell. In the present study, a novel mathematical zero-dimensional model has been formulated for the water mass balance and hydration of a polymer electrolyte membrane. This model incorporates all the essential fundamental physical and electrochemical processes occurring in the membrane electrolyte and considers the water adsorption/desorption phenomena in the membrane. The effect of diffusivity model, surface roughness and water content driving force is considered. The model is validated against experimental data. In the results it is shown that the fuel cell water balance calculated by this model display better fit with experimental data-points compared with model where only water transport equilibrium were considered. We conclude that this discrepancy is due a different rate of water transport when membrane absorption/ desorption is considered in the model. Many studies have erroneously considered fuel cell membrane in equilibrium condition disrigarding non-ficknian behavior of the cell membrane. For this reason the model presented in this study becomes useful when studying PEM fuel cell system in dynamic conditions.

A Novel Model of Water Hydration in PEM Fuel Cell in Dynamic Operations

Polymer membrane (PEM) fuel cell requires good hydration in order to deliver high performance and ensure long life operation. Water is essential for membrane proton conductivity which increases by nearly six orders of magnitude from dry to fully hydrated. Adequate water management in PEM fuel cell is crucial in order to avoid an imbalance between water production and water removal from the fuel cell. In the present study, a novel mathematical zero-dimensional model has been formulated for the water mass balance and hydration of a polymer electrolyte membrane. This model incorporates all the essential fundamental physical and electrochemical processes occurring in the membrane electrolyte and considers the water adsorption/desorption phenomena in the membrane. The effect of diffusivity model, surface roughness and water content driving force is considered. The model is validated against available experimental data. In the results it is shown that the fuel cell water balance calculated by this model display better fit with experimental data-points compared with model where only water transport equilibrium were considered. We conclude that this discrepancy is due a different rate of water transport when membrane absorption/ desorption is considered in the model. The model becomes useful when studying fuel cells in dynamic conditions.

Correlation of In Situ and Ex Situ Measurements of Water Permeation Through Nafion NRE211 Proton Exchange Membranes

Water permeability at is determined for Nafion NRE211 membrane exposed to either liquid or vapor phases of water. Chemical potential gradients of water across the membrane are controlled through use of differential humidity (38–100% RH) in the case of water vapor and hydraulic pressure in the case of liquid water. Accordingly, three types of water permeation are defined: vapor-vapor permeation, liquid-vapor permeation (LVP), and liquid-liquid permeation. The difference in chemical potentials across the membrane, and more significantly, the flux of water, is largest when the membrane is exposed to liquid on one side and vapor on the other (i.e., LVP conditions). Polarization curves and net water fluxes are reported for NRE211-based MEAs at under two different operating conditions. Water permeability measurements obtained ex situ are compared to fuel cell water balance measurements obtained in situ. It is found that the magnitude of back-transport of water during fuel cell operation can be explained only by considering that the membrane is exposed to liquid on one side and vapor on the other (i.e., LVP conditions). Thus, LVP water transport is largely responsible for regulating water balance within the operating membrane electrode assembly.

Water Management in Cathode Catalyst Layers of PEM Fuel Cells: A Structure-Based Model

The cathode catalyst layer (CCL) is the major competitive ground for electrochemical reaction, reactant transport, and water and heat exchange in a polymer electrolyte fuel cell (PEFC). Nevertheless, it is often treated as a thin interface. Its pivotal role in the fuel cell water balance is unexplored. Here, the structural picture of CCLs forms the basis for a novel model that links spatial distributions of processes with water handling capabilities and current voltage performance. In the first step, the statistical theory of random composite media is used to relate composition, porous structure, wetting properties, and partial saturation to effective properties. In the second step, these effective properties are used in a macrohomogeneous model of CCL performance. A set of reasonable simplifications leads to a full analytical solution. Results demonstrate that the CCL acts like a watershed in the fuel cell, regulating the balance of opposite water fluxes toward membrane and cathode outlet. Due to a benign porous structure, the CCL represents the prime component for the conversion of liquid to vapor fluxes in PEFCs. Furthermore, the CCL is highlighted as a critical component in view of excessive flooding that could give rise to limiting current behavior.