Patent Application
20170141416
The present invention relates to a novel internal humidification technique for low temperature proton exchange membrane fuel cell (LT-PEMFC). More specifically, the invention relates to a wick based passive technique for supply of water directly to the membrane electrode assembly by capillary action which render simultaneous humidification and cooling of PEM fuel cells, with high electrical efficiency.
The proton exchange membrane used widely today by most PEMFC stack manufacturers is based on polyflouroethylenesulfonated polymer which requires sufficient levels of moisture content for effective proton transfer from anode to the cathode. Thus, it is needed that the gases supplied to the PEMFC are humidified either externally or internally to maintain the required moisture content in the membrane and to get continuous power output. In conventional systems, humidification of gas streams is achieved through external humidification unit comprising of a membrane (Nafion based) humidifier and water supply.
In such a humidifier, liquid water and dry air are fed to either side of the membrane thereby creating a water activity gradient across the membrane. This forces the water to diffuse across the membrane from the water side to the gas side that subsequently evaporates at the membrane/gas interface resulting in humidification of gas. The external humidification system adds to the parasitic power loss apart from adding to the system complexity. The use of humidified gas streams is very critical on both the electrodes. At anode typically, more humidification is required (>80%) as the electro-osmotic drag leads to more dehydration at the anode side of the membrane. At cathode, water is produced because of the reaction between the protons and the incoming oxygen. However, the compressed air that is supplied to the cathode is too dry for maintaining proper moisture levels. Thus, some amount of humidification (˜50%) is required for the cathode side stream as well.
More lucrative way of achieving humidification is internally. By supplying water directly to the gas streams in the form of mist (by means of ultrasonic nebulization) or by using water transport plates as bipolar plates for gas flow, one can achieve internal humidification. Former is energy intensive while use of latter normally results in less fuel utilization due to fuel crossover through the plates. In the present method, water is brought in contact directly with MEA by passive wicking action.
U.S. Pat. No. 6,960,404 describes the wicking action over the cathode in the form of channels in order to supply the water inside the cell for internal humidification. The wicking material should be porous and hydrophilic although it does not need to be electrically conducting. However, the drawbacks of such a configuration of fuel cell is that it is difficult to fabricate bipolar plates with wicking materials as channels and also usage of electrically insulating wicking material shall render increased resistance to charge transfer across the cells in a fuel cell stack.
EP 1,530,813 describes a technique for direct supply of water to the bipolar plate channels for internal humidification of membrane. The technique uses an extremely thin metal foil (˜40 microns) assembly designed for placing over the top and bottom portions of active area of the flow field plate. The water is supplied to the metal foil directly through a complicated design of bipolar plate. The water when reaches the flow field evaporates into the gas stream thereby humidifying it. The gas that is reacted humidifies the membrane, and the remaining gas is vented to provide cooling. The method though effective, does not alleviate the problem of easy manufacturing globally and works only with metallic plates.
EP 2,016,640 B1 describes a novel bipolar plate design for direct water supply to the metal foil as an extension of their previous patent (EP 1,530,813 B1). The bipolar plate design provides an easy transport of water into the cell but is still quite complicated as it requires significant maneuvering in the bipolar plate design which incurs increment in manufacturing cost.
In another form of wicking technique used for In-Situ humidification in PEMFC is shown in U.S. 2004/0170878, which describes porous fibers that can be used as channels for direct water transport into the cell which is briefly illustrated in
The wicking evaporative cooling described in the aforementioned publication is stated to require external water, from a source outside the fuel cell power plant, since the water generated at the cathode (process water) is said to be insufficient, except at startup, to achieve the necessary cooling. This is also true in an evaporatively cooled fuel cell stack which relies on wicking as in U.S. Pat. No. 4,826,741. Therein, 100 cm2 cells have performance of only 0.7-0.8 V at 100-120 mA/cm2 (108-130 A/ft2).
In view of the aforesaid problems of prior art, current work was undertaken by the inventors to propose use of an electrically conducting & porous material for its ability to provide direct membrane humidification using several configurations of wick based 100 cm2 PEM fuel cell setup. The same was later extended to a 5-cell PEM stack to check for the scalability of this technique and was found to be performing with high efficiency.
The main object of the present invention is to provide a method of retaining the water content in the membrane during the operation of fuel cell by developing a cost-effective humidification technique in Low temperature Proton Exchange Membrane Fuel Cell (LT-PEMFC), that would be easy to scale up and avoid system complexity.
Another object of the present invention is to provide a passive means of water transport into the cell thereby reducing the parasitic power loss which is normally incurred by conventional fuel cell systems.
Still another objective of the present invention is to increase the hydrophilic nature of the wicking material by treatment with acetone or Tin oxide.
Yet another objective of the present invention is to provide a water supply process in the form of a trough that should be placed at the bottom of the fuel cell/stack.
In still other objective, the present invention would facilitate cooling of fuel cell stack in operation thereby reducing the parasitic power loss incurred by the conventional cooling systems.
The present invention deals with electrically conducting, porous and hydrophilic layer of thickness less than 1 mm being used as a wicking material for water management in PEMFC. The invention therefore provides for water to be sorbed through capillary action in the wicking material. The presence of wicking material adjacent to the MEA facilitates for the moisture content required by the membrane for effective proton transport from anode to cathode.
Accordingly, the present invention provides a fuel cell setup comprising the wicking material placed over the gas diffusion layer. The dry gas while passing through the flow field takes up the water from the porous wick surface and gets humidified. It is further advantageous in some fuel cell configurations to have the gas flow in the direction of capillary action (co-flow mode) to bolster the water transport up the wick. The technique reported in this invention exhibited improved cell performance in comparison with conventional humidification technique under cell temperatures of around 55-60° C. The technique when tested on a 5-cell stack worked as good as conventional fuel cell system when the wick was placed over the cathode. Thermal management was more incumbent in case of stack.
The technique of the invention facilitates cooling with stable performance under constant load conditions. Thermal management studies back up the results.
Efficiency of stack as per the invention was found to improve with inlet down configuration resulting in cooling of 16.5 W at 500 mA/cm2 current density.
Overall cooling achieved as per the invention was 8.15% out of the total 202.5 W available heat.
Maximum cooling of approximately 11% is achieved at a current density of 300 mA/cm2 with co-flow configuration. Even with this much amount of cooling, net parasitic load could be reduced considerably from the cooling circuit.
For a better understanding of the present invention, and to show more clearly how it may be carried out into effect, reference will now be made by the accompanying Figures, which show the preferred embodiments of the present invention.
In order to achieve the objectives of the current invention, the inventors have provided a novel approach for cost-effective in situ humidification of the membrane in LT-PEMFC.
According to the invention, the technique makes use of principles of evaporative cooling, capillary action in electrically conducting, porous and hydrophilic layer of size less than 1 mm to supply the water directly into the cell for humidification of membrane internally.
In accordance with an embodiment of the invention, the hydrophilic character of the wicking material is further increased by treatment with acetone or Tin oxide dispersion treatment.
It is known that for optimal performance of fuel cell, water management in crucial.
The membrane should neither get dried up due to electro-osmotic drag nor should the cathode get flooded due to excessive water production at high current densities. Flooding is usually not a problem on the anode except for operation at lower current densities. In general, auxiliary components like humidifiers and pumps are used for achieving the required humidity Of reacting gas streams resulting in increased parasitic power loss, increased cost and complexity of system.
The present invention aims .at providing an easy to scale up passive technique for humidification in LT-PEMFC while avoiding introducing any complexity to the existing design of proton exchange membrane (PEM) fuel cells, while providing overall high efficiency also. Almost all of the techniques being used in the prior art address water management at the cathode with or without dry feed operation and do not consider water management over the anode using such methodologies. Moreover, inventors extend the fuel cell area to 100 cm2, which has not yet been studied with existing wick-based techniques. Therefore, the present study deals with the use of an electrically conducting & porous material for its ability to provide direct membrane humidification using several configurations of wick based 100 cm2 PEM fuel cell set-up. According to the current invention as in
According to the invention, the wicking material has pore size in accordance to the invention is in range of 25-94 μm and size of 0.2-1 mm.
According to another embodiment of the invention the wicking material has thermal conductivity of more than 2 W/m.K and electrical resistivity <500 mOhm-cm.
According to the invention the electrically conducting film includes but is not limited to carbon cloth and metal foam sheets. The carbon cloth in its dry form is more permeable to gas transport while in wet form offers increased resistance to the gas transport. Despite then, no reduction in performance of the fuel cell has been observed.
According to the invention the average contact angle of the wicking material/carbon cloth is 77 degrees.
According to the current invention as in
According to the current invention, a trough is used for water transport up the wicking material used. It is important that no two wicks come in physical contact with each other in order to prevent short circuiting. According to the invention, the trough has been designed to provide separate compartments for water transport for each electrode in the fuel cell.
The trough design should comply with the water requirements for consistent and stable fuel cell performance. Thus, the trough should be designed in order to accommodate a minimum quantity of water at all times depending of fuel cell power. The height and compartments dimension can be adjusted accordingly. It should be noted that the portion of the wick which is out of the fuel cell needs to be dipped as much in water as possible for better wicking action. It is important that the material chosen for fabrication of trough is electrically insulating and can hold sufficient amount of water in order to operate the fuel cell for longer duration without the need of refill.
In accordance with the embodiments of the invention, de-ionized water has been used for water transport as it offers reduced electrical conductivity and is important to prevent transfer of ionic impurities into the cell for better durability of MEAs. The flow rate of reactant gases to be used should take into account the differential capillary pressure and the wicking velocity. The stoichiometry should be considered to provide the maximum wicking velocity for a particular wicking material.
According to the invention, the gas can be supplied either in counter-flow mode or co-flow mode in order to facilitate or prevent wicking action. In configurations wherein only anode or cathode is being internally humidified, the other electrode is supplied with the conventional counter-flow mode.
According to the invention, hydrogen is supplied as shown in
According to the present invention, there may be four different configurations of a PEMFC system depending on the placement of the wicking material on cathode or anode or both. It is known that humidification in PEMFC is more critical over the anode because of electro-osmotic drag. However, small amount of humidification (˜50% RH) is still necessary on the cathode in order to avoid much reduction in partial pressure of oxygen. Thus, the present invention was tried with four different configurations. The configuration without the wick was used as control. In other configurations, the wick was placed over the cathode or anode or both and was dipped in the trough placed at the bottom of the fuel cell setup. The respective electrode(s) with wick were fed with dry gas with 0% RH for fuel cell operation and the performance was recorded using the polarization technique. All the experiments were performed at ambient pressure and 55° C. on a serpentine flow field based fuel cell setup.
Electrically conducting hydrophilic layer as used in the invention has adhesive interactions with water molecules. The adhesive forces have to be greater than the cohesive forces between different water molecules for them to produce capillary action. In these terms, the contact angle determines which force is dominant Lower the contact angle more is the wetting energy and so is the adhesive force thereby facilitating the capillary action. Water molecules are absorbed through the voids of the cloth so that when the dry gas passes through it, it takes with it the required moisture for the membrane.
While the above description constitutes the preferred embodiments, it is noteworthy that the present invention is susceptible to modifications and changes without departing from the fair meaning of the proper scope of the accompanying claims. For example, a wide variety of electrically conducting and hydrophilic fabrics or extremely thin metal foam meshes can be employed. As will be appreciated by those skilled in the art, the requirement for humidification is dependent on the type of wicking material used, its pore size distribution, electrical conductivity and hydrophilic nature.
The novelty of the invention lies in the concept of wick-based humidification is derived based on the principle of evaporative cooling, which can render simultaneous humidification, and cooling of proton exchange membrane (PEM) fuel cells. Entire process is psychrometry of air-water system. The work encompasses studies at two phases, one at single cell level and then at 5-cell stack level. The single cell studies were performed with all the possible configurations that one can have for this technique. The water produced at the cathode is found to be sufficient for stable flood-free performance under dry air feed in configuration-C. Compact arrangement of trough in the present study can also facilitate wick based humidification at the anode (configuration-B) giving a peak power density of 409 mW/cm2 in co-flow mode which is nearly equal to the peak power density of 408 mW/cm2 in case of control. Configuration-D does not show performance comparable to configuration-A in both the flow modes. Temperature and voltage profiles for the configuration-C in a 5-cell stack were obtained and compared with the control. The maximum cooling that could be achieved was close to 16.5 W for a total available heat of 202.5 W
The following examples are given by way of illustration and therefore should not be construed to limit the scope of the present invention.
Pore size distribution, gas permeability and contact angle play an important role for wicking action to take place in a porous hydrophilic material. It also gives an idea about the possibility of flooding of electrodes due to excess wicking. In order for wick to supply water to the fuel streams internally, it is essential that a positive wicking velocity be continuously maintained. Hence, optimum pore size distribution and permeability are critical for flood-free cell performance. Wicking action is in turn determined by using two important factors namely differential capillary pressure and wicking velocity. Differential capillary pressure (dPcap) is a function of surface tension of water, contact angle and pore diameter, whereas wicking velocity depends on the difference between dPcap and differential flow field pressure (dPflowfield), permeability, viscosity of fluid and channel length. Average pore size, air flow rate and permeability through the wick were measured in relation to varying differential air pressure. Sessile drop method was used to measure the contact angle over the wick surface. Thermal stability of the wicking material was studied for a temperature range of 30° C. to 250° C. using a Thermo-gravimetric analyzer (TGA).
It is important that the wicking material used for water transport be thermally stable for the PEMFC temperature operation range. In general, wicking materials like cotton loose about 5-7% of their mass in the temperature range of 30-250° C. Hence, long duration operation with such wicking materials is not likely to provide stable performance. From
Configuration-A corresponding to the control single cell setup with external humidification of gases was taken as the base line performance for comparison with other configurations which make use of wick for internal humidification. The cell was supplied with 80% RH at the anode and 50% RH at the cathode. The humidification levels were maintained by controlling the humidifier temperature and gas line temperature. The gas line temperature was maintained at 60° C. for both anode and cathode. The humidifier temperature was maintained at 50° C. for anode and 45° C. for cathode. Galvanostatic polarization curve was recorded after cell stabilization for 6-8 h. The stoichiometry used for all tests was 1.2 for anode and 3 for cathode.
Anode Side Internal Humidification Experiment with Cathode Side Externally Humidified (Configuration-B)
Configuration-B corresponds to the single cell setup with the wicking material over the anode of the MEA. The wicking fabric extends out from the bottom of the cell and dips inside the trough containing water. The fabric starts distributing the water through its voids by capillary action. The anode is operated with counter-flow and co-flow modes in the fuel cell. The cathode is supplied with gas only in counter-flow mode. In this configuration, the fuel cell was supplied with 50% humidity at the cathode while the anode was supplied with dry stream of hydrogen gas (0% RH). Galvanostatic polarization curve was recorded after cell stabilization for 6-8 h. The stoichiometry used for all tests was 1.2 for anode and 3 for cathode.
Cathode Side Internal Humidification Experiment with Anode Side Externally Humidified (Configuration-C)
Configuration-C corresponds to the single cell setup with the wicking material over the cathode of the MEA. The wicking fabric extends out from the bottom of the cell and dips inside the trough containing water. The fabric starts distributing the water through its voids by capillary action. In this case, cathode is operated with counter-flow and co-flow modes while the anode is operated only with counter-flow mode. In this configuration, the fuel cell was supplied with >80% humidity at the anode while the cathode was supplied with dry stream of air (0% RH). Galvanostatic polarization curve was recorded after cell stabilization for 6-8 h. The stoichiometry used for all tests was 1.2 for anode and 3 for cathode.
Configuration-D corresponds to the single cell setup with the wicking material over both sides of the MEA. In this configuration the cell was dry fed with anode and cathode gas streams and was allowed to provide internal humidification through wicking action. Galvanostatic polarization curve was recorded after cell stabilization for 6-8 h. The stoichiometry used for all tests was 1.2 for anode and 3 for cathode.
Thermal Management Experiment with Configuration-C Used in a 5 Cell Stack
As an extension to the capability of this technique, thermal management studies were conducted to study the performance with configuration-C using a 5 cell stack. The stack was built with wick placed over the cathode for all 5 cells. The stack was fed with dry air (0% RH) at the cathode and >85% RH over the anode. The stoichiometry was kept constant for all tests at 1.2 for hydrogen and 3 for air. Temperatures were measured at the two end plates using thermocouples. The temperature and voltage profile were measured with increase in time. The test was terminated beyond 50° C. end plate temperature. The tests were repeated for multiple load conditions of 200-500 mA/cm2. Finally polarization curves were obtained, with cut off voltage being 1 V. Finally, the results obtained for configuration-C in co-flow and counter-flow modes were compared with the control setup (without cooling).
In situ humidification experiments were carried out on a 100 cm2 single cell setup. Hydrogen Screener Membrane Electrode Assembly (MEA-5 layer), Active Area 100 cm2 was procured from Alfa Aesar, Johnsan Mattey Pvt. Ltd., India. AvCarb carbon cloth procured from Nikunj Exim Pvt. Ltd. was used for all the tests. All the tests were conducted on BioLogic fuel cell test station with a load capacity of 140 A and voltage range of 0-5 V.
Apparatus for single cell experiments include:
All the above mentioned components together make up the complete setup for single cell experiments.
Based on the configurations, experiments were carried out for a 100 cm2 single cell keeping all the parameters constant except for the placement of wicking material and flow of gases in the co-flow or counter-flow scenario. The studies were systematically conducted wherein at first control (Configuration-A) setup was tested with all conventional units and was taken as benchmark for comparison with other configurations (B, C and D). Thus, any deviation from the performance obtained in the control experiment was attributed to be (i) due to the presence of cloth over the MEA (ii) dry and ambient feed of gases (iii) water wicked at ambient temperature and (iv) change in flow mode of gases. The performance in any case was quantified by means of a polarization curve and the effect of adding the cloth was tested by means of electrochemical impedance spectroscopy.
The polarization data for the single cell was obtained at 55° C. at varying load current densities with hydrogen stoichiometry of 1.2 and air stoichiometry of 3, set for a current density of 0.8 A/cm2. The tests were carried out using counter-flow and co-flow modes.
Small amount of water was wicked up in case of Configuration-B even under the counter-flow mode where the maximum power density was restricted to only 344 mW/cm2. This was mainly due to increased gas diffusion resistance provided by the wet wick and can also be attributed to the dry feed of hydrogen which can lead to increase in membrane resistance despite the wicking action in counter-flow mode. In case of configuration-C, the maximum power density was measured at 465 mW/cm2. This was expected, as the air stoichiometry was kept sufficiently high which would prevent any significant gas diffusion resistance and the water produced at the cathode was more uniformly distributed by the air flow over the wick thereby preventing any flooding events over the cathode.
Electrochemical impedance measurement studies were conducted after every polarization study. All the parameters were kept constant as were used for polarization studies. Only the program in BioLogic fuel cell test station was changed to perform impedance spectroscopic analysis for the cell. Frequency range was set at 0.1 Hz to 10 kHz and voltage was kept at 0.6 V which lies in the normal operating voltage range of fuel cell.
Nyquist plot was then recorded within the frequency range for every configuration.
Impedance spectroscopy is a powerful tool for understanding the behavior of an electrochemical system. In general Nyquist and Bode plots are used to give the complete information for any impedance spectroscopy measurements. Where Nyquist plot gives the details of the real and imaginary parts of overall impedance, Bode plot gives the system's frequency response.
For the co-flow mode,
Thermal management is an important aspect of this study. It is confirmed from the single cell humidification experiments that the technique is capable of providing in situ humidification in all configurations and even better performance as compared to the conventional cell setup with configurations B and C. Therefore, for thermal management studies which were conducted with a 5-cell stack, only configuration A and C were tested to establish a valid comparison between a noncooled stack (configuration A) and stack with wick at cathode (configuration-C). The tests were conducted with both counter and co-flow modes. It was therefore established by means of temperature and voltage profile that the stack with wick had better thermal management ability as compared to non-cooled stack.
The stack based on configuration-C was built by placing carbon cloth carefully in between every cathode of the MEAs and flow field plate within the 5-cell assembly. The stack was built using four bipolar and two monopolar plates with serpentine flow field. Five new 100 cm2 MEAs were used for thermal management experiments. All the graphite plates used were procured from M/s Schunk Kohlenstofftechnik GmbH, Germany. No cooling plates were added into the stack. All the remaining components were kept same as in the case of single cell experiments.
In order to determine that cooling has occurred using the proposed technique, it is imperative to determine the temperature profile of the stack with time under a given load condition. In order to simulate real time practical conditions, the tests were conducted under ambient conditions with room temperature close to 30° C. All the ceiling fans in the testing room were switched off during the experiments in order to avoid major forced convective heat transfer from the periphery of stack to the atmosphere. Tests were conducted for configuration A and C only. Stack was operated under constant load conditions of 20, 30, 40 and 50 A respectively, with anode side humidification kept above 80% at all times for both the flow modes. The load was given by means of a load box (Bitrode LCN power module). Cathode was fed with dry air (0% RH) for configuration-C and with 50% RH for configuration A. The humidification setup for stack comprises of two water baths, two pumps and two PermaPure humidifiers (model FC 300-1660-10LP). For control, the water bath temperatures were set at 55° C. for anode humidification and 45° C. for cathode humidification. The water from the baths was circulated into the humidifier by means of peristaltic pumps. The humidifiers were fed with counter flow of gas and water. The gas outlet of humidifier was connected to the inlet gas port of the fuel cell stack. Since, the gas line was open to atmosphere; the gas outlet of humidifier was kept close to the inlet of stack in order to prevent any loss in water vapor in the condensation process. In case of configuration-C, a humidity sensor was attached to the cathode outlet in order to quantify the relative humidity and temperature of the dry gas after passing through the flow field. Temperature was measured at both the end plates by means of two thermocouples and was recorded manually at every 20 seconds interval. Stack temperature was taken as mean of the two temperatures. The tests were terminated once the end plate temperatures reached 50° C. However, it has to be understood that the actual cell temperature over the active areas of the MEA can be 5-7° C. higher than the temperature measured at the end plates.
Temperature profile of the stack was measured at the two end plates. The temperature profile of the stack measured between room temperature and 50° C. would give a clear perspective on cooling ability of the stack. The pattern of rise in temperature at different load conditions give a clear idea of the amount of cooling that is achieved through incorporation of wick at the cathode.
The experimental setup and different parameters were kept similar to the ones kept for temperature profile study. In these experiments, stack voltage was recorded for the stack temperature rise from room temperature to 50° C. The voltages were recorded for load conditions (20, 30, 40 and 50 A) similar to the temperature profile experiments. The purpose of this experiment is to establish the stability of voltage with rise in temperature of fuel cell stack.
It is important that the performance is maintained while the stack reaches its operating temperature. Since, stack was operated at ambient pressures, the stack temperature was kept below 60° C. Voltage profile gives a clear picture of the stack performance with increase in temperature. A stable voltage for longer duration of operation indicates a better performance.
Polarization curves for stack configurations were obtained by recording the stack voltage for increasing current density. The current density was increased in steps of 10 seconds each starting from 5 mA/cm2 to 600 mA/cm2. The range was divided into 40 steps wherein the voltage was recorded at each step for 10 seconds. Cut off voltage for termination of test was set at 1 V. Since, the stack had been operated for several hours during temperature and voltage profile studies polarization curve was obtained as soon the stack reached near 50° C. temperature All the tests were conducted at ambient pressure conditions. Polarization curve was obtained for a stack temperature range of 50-55° C. with hydrogen and air stoichiometry at 1.2 and 3 respectively.
Efficiency of the stack has been calculated based on the theoretical potential obtained from the lower heating value of change in enthalpy for hydrogen-oxygen reaction. Therefore the efficiency of the fuel cell stack can be obtained by a simple formula:
Fuel cell output power is obtained by the following relation:
P
o(Watts)=Current density (A/cm2)×Voltage (V)×Active Area(cm2)
Waste heat then can be calculated by the relation:
P
H (Watts)=J (A/cm2)×(Vth−V)×No. of cells×Active area (cm2)
Theoretical output power is actually the sum of waste heat and fuel cell power output. It can also be calculated by a simple equation as mentioned below:
P
th (Watts)=Vth×J (A/cm2)×No. of cells×Active area(cm2)
Table 2 documents the efficiency of various configurations studied during the thermal management experiments. It can be clearly observed that the heat generation rate increases with increase in current density. However, the efficiency goes down due to lower power output.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1797/DEL/2014 | Jul 2014 | IN | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IN2015/050056 | 7/2/2015 | WO | 00 |
