A Review of Experiments on Cold Start of PEM Fuel Cells

Azizul bin Mohamad1, a 1 PPK Mekatronik, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysia

a azizul@unimap.edu.my

Keywords: PEM fuel cell, sub-freezing start up, cold start, experimental studies

Abstract. This paper evaluate previous experimental studies on sub-freezing start up of proton

exchange membrane (PEM) fuel cell system, and identify issues for further investigation. In a

successful cold start, product water from electrochemical reaction in the cathode must be removed

from the cell before it turns into ice and causing voltage drop and shutdown also leads to permanent

damage to fuel cell components. Successful single PEM fuel cell start up was achieved from

temperature as low as -30°C. Some researchers found that cold start of a 30 W stack from -20°C

was possible only with aid of external energy. Successful self start up a 2 kW stack from

temperature -5°C was reported but the time taken was unacceptably long and attempts to start up

the stack at lower temperatures were failed. Based on the current state of research, further research

is necessary to fully understand the operation and mechanism of PEM fuel cell cold start.

Introduction

Fuel cell technology is becoming more and more important in recent years due to its promising

prospect as an alternative to fossil fuel based power system and its positive impact to the

environment since no or minimal pollutant is released to the environment during the operation of

fuel cells [1]. Despite these significant benefits, several major obstacles need to be overcome before

fuel cells could be fully commercialized. The obstacles exist in every aspects of fuel cell life,

ranging from hydrogen production, storage and distribution, cost and material suitability, to the

performance and durability of fuel cell system at various applications and operating condition [2].

Heat is a by-product of electrochemical reaction in the fuel cell. The heat is generated due to

combination of stack inefficiencies, water management inefficiencies in term of condensate water

recovery, as well as balance of plant inefficiencies due to compressors, motors and drive train

losses. During cold start there is also heat released due to phase transition of water from solid into

liquid. The handling of these heat loads is commonly termed as “thermal management”. At above-

zero operation, this waste heat must be removed from the fuel cell using passive and active cooling

methods in order to maintain the cell operating temperature at its optimum range. On the other

hand, at sub-zero temperatures, the waste heat is used to warm up water, another product of fuel cell

electrochemical reaction, in order to avoid liquid water from freezing.

The behavior and performance of proton exchange membrane (PEM) fuel cell in automotive

applications at sub-freezing temperatures is one of major issues that need to fully resolved and

understood before fuel cell vehicles could be commercialized and mass produced. At sub-freezing

temperatures, product water from electrochemical reaction in the fuel cell may freeze thus prevent

further reaction take place and may cause permanent damage to the fuel cell components. In recent

studies, various researchers [3, 4] had conducted investigations on this matter via experimental and

modeling works, but still a significant issues need to be fully resolved before a prominent solution

can be obtained. This paper evaluates previous experimental studies on sub-freezing start up of

proton exchange membrane (PEM) fuel cell system, and identifies issues for further investigation.

Experimental Work on Cold Start of PEM Fuel Cells

One of the main issues of PEM fuel cell operation at subzero temperatures is cold start ability.

The Department of Energy, United States of America set technical target that by 2010, PEM fuel

cell stacks will be able to start up from -20°C to maximum power in 30 seconds [2]. In a successful

cold start, product water from electrochemical reaction in the cathode must be removed from the

Applied Mechanics and Materials Vol. 315 (2013) pp 851-855Online available since 2013/Apr/10 at www.scientific.net© (2013) Trans Tech Publications, Switzerlanddoi:10.4028/www.scientific.net/AMM.315.851

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP,www.ttp.net. (ID: 130.207.50.37, Georgia Tech Library, Atlanta, USA-13/11/14,13:47:48)

cell before it turns into ice and causing voltage drop and shutdown as well as leads to permanent

damage to fuel cell components. Ideally, heat generated by the electrochemical reaction should be

sufficient to maintain product water in liquid phase while at the same time elevate the temperature

of the cell to above freezing, thus eliminating the needs for external heating.

Hishinuma et al.[5] investigated the performance of a PEM fuel cell start up at a temperature

below freezing by using a single cell PEM fuel cell with a 104 cm2 effective area of membrane and

electrode assembly in their constant voltage experiments as well as in their numerical model. They

kept the gas utilization constant by varying the gas flow rate with changes of current for test

temperature of 40°C to -25°C and pressure 1 to 2 atm. They noted the decrease in PEMFC start up

performance at higher current densities and pressures, as well as at lower cell temperatures, due to

higher production of ice on the reactive area of the cathode. They adjusted the current densities and

gas flow rate to balance the rate of water production and removal in order to maintain the cell

performance at freezing temperatures. They concluded that at temperature -5°C, heat generated in

the fuel cell is sufficient to warm the fuel cell and enable self-starting while for any temperature

below that, an additional heat source need to be supplied to the fuel cell to enable start up. Oszcipok

et al. [6] investigated the cold start performance of PEM fuel cell single cells as well as portable

PEM fuel cell stacks. For single cells, the experiments were conducted in potentiostatic and

isothermal mode. The experimental parameters were membrane thickness, types of gas diffusion

layer, gas flow rates and levels of membrane humidification. They utilized statistical method to

analyze the result and obtained positive correlations between the dry membranes and high air flow

rates with the cold start. They then conducted further experiments on six-cell stack in galvanostatic

mode. The experimental parameters were stack impedance, load current and temperature. They

concluded that it was possible to achieve cold start without any external heating from temperature

as low as -10°C. They also found that cell start up was strongly affected by the stack impedance. In

another literature [7], they studied various key design parameters of a 30-W portable fuel cell

system operating between -20°C to 40°C with the aid of a thermal model and concluded that active

heating system is required to achieve successful cold start of a portable PEM fuel cell system since

frozen water may cause irrecoverable damages to the fuel cell and active heating system may

consume less energy that its passive counterpart. They reported a successful cold start and extended

operation of a 30 W portable PEM fuel cell system at -20°C using electrical heating to assist start

up.

Hottinen et al. [8] examined the effect of subzero temperatures on constant current density

operation as well as cold start ability of planar free-breathing PEM fuel cell. For constant current

density operation, the temperature varied from 0 to -27.5°C while for cold start operation, the

ambient temperature was set at -5°C and -10°C. They found that the fuel cell able to operate with

stable performance at higher current densities since the heat generation was sufficient to prevent

water freezing inside the cell. They also noted the ice formation outside the cathode side. For cold

start performance, they found that at -5°C, the cell with dry membrane was able to start without any

performance loss. The cell also able to start at -10°C with minor performance loss provided the

starting procedure was slow enough. Yan et al. [9] investigated the influence of sub-freezing

temperature on a 25-cm2 PEM fuel cell performance, start up and fuel cell components. They found

similar result as other researchers that pre-purged; insulated PEM fuel cell was capable to start up at

-5°C without any deterioration in performance. The cell was able to operate up to -15°C but

irreversible performance loss occurred if the cathode operating temperature drops to lower than -

5°C. They used scanning electron microscope (SEM) to analyze the cell after subfreezing operation

and found severe damage to the membrane electrode assembly and backing layer. Tajiri et al. [10]

utilized a newly developed experimental protocol, isothermal cold start, in their experiments to

elucidate the basic physics of fuel cell cold start. Isothermal cold start was achieved by fixing the

cell temperature constant at the start up ambient temperature using single cells with high thermal

mass. It was used to investigate the intrinsic cold start capability of membrane-electrode assembly.

They proposed a method of equilibrium purge using partially humidified gas with controllable level

of relative humidity was suggested to manage the distribution of water inside a cell prior to cold

852 Mechanical & Manufacturing Engineering

start up. They also used dry purge method in their experiments which was a realistic simulation of

fuel cell vehicle operation where the initial membrane water content was controlled by purge

duration. They concluded that the equilibrium purge could effectively maintain the cell internal

condition prior to cold start, which then lead to higher consistency and reproducibility of the

PEMFC cold start experiments. Tajiri et al. [11] also studied the effects of startup temperature,

current density, and the membrane thickness on the PEMFC cold start capability. They adopted the

amount of product water in mg/cm2 during start up as an index to quantify the cold start capability.

They reported that the startup temperature strongly affects cold start performance of PEMFC and at

-3°C the cell can operate for an indefinite period. They concluded that the startup current density

strongly influenced the water production for both purging methods, with the higher production of

water at higher current density since lesser time was offered for the membrane to absorb and stored

the water produced. The results and the analysis are in accordance with Oszcipok.

Pinton et al. [12] studied the cold start behavior of 220 cm2 PEM single cells using isothermal

potentiostatic and galvanostatic tests for different parameters, namely the initial membrane water

content, the operating voltage, the cell temperature and the current. They found that the optimal

level of fuel cell core wetting occurred when the cumulative heat generation in the electrochemical

reaction is maximal. They deduced the overall fuel cell performance evolution from membrane

water management analysis and cell resistance measurements. They suggested that the reduction in

fuel cell performance in term of fuel cell starvation were caused by the ice formation in the cathode

layer pores which inhibit oxygen transport and by the ice formation in active region sites which

increase the electrical resistance of the cell. They plotted characteristic curves after each shutdown

and start up at freezing temperature and observed that the performance degradation of fuel cell was

less than 1% per cold start at rated conditions. They also concluded that in order to achieve self start

up within acceptable duration of automotive applications (<30 s), low value of initial voltage (0.3 <

Ucell < 0.5 V) should be used.

Bégot et al. [13] presented the design and validation of a 2 kW fuel cell test bench for

subfreezing studies. This test bench was devised to evaluate the effects of ambient temperature, gas

and coolant flow rates, current density as well as fuel cell impedance on the cold start performance

of PEM fuel cell. The experimental setup was developed to emulate a fuel cell system in parked

vehicle in a freezing environment. The fuel cell, its coolant circuit and main sensors are placed

inside a climatic chamber while the main part of the test bench is at normal temperature. In another

literature, Bégot et al.[14] used their test bench to evaluate the influence of current density, pre-start

stack impedance at 1 kHz, gas flow rate, gas pressure, coolant flow rate and ambient subfreezing

temperature on a 2 kW PEMFC cold performance. Based on the outcomes of the experiments, they

established that the best cold start performance could be achieved by using a combination of low

current density, high pre-start impedance, moderate subfreezing temperature (-5°C), high gas flow

rate, low gas pressure and low coolant flow rate. Using these parameters, they found that the self

start up of fuel cell was achieved at -5°C in 30 minutes with no aid of external energy, while both

self start up attempted at -10°C and -15°C were not successful. This result showed that the self start

up of a 2 kW PEM fuel cell stack from -5°C would take a relatively longer time than single cells

and it may not succeed at much lower temperatures. From the results, they suggested the existence

of three distinct phases on freeze mechanism: (a) first phase, transient phase of membrane

humidification due to dry membranes and low current; (b) second phase, ice clogging occurred on

the active layers; and (c) third phase, a variable quantity of the produced water arrived at the gas

diffusion layers and channels.

On the other hand, Jiang and Wang [15] explored the potentiostatic start up of PEM fuel cell

from subfreezing temperatures. In this method, the cell voltage is fixed while the current density is

allowed to vary. The current density was reported to increase significantly during the cold start

procedure due to membrane hydration and increase in cell temperature. This amplification promotes

more heat generation and expedites the temperature elevation in the fuel cell. They pointed out that

the only situation where the potentiostatic start up is better than galvanostatic (constant current

density) start up is when the membrane is dry following gas purging procedure. Nevertheless, these

Applied Mechanics and Materials Vol. 315 853

researchers proved numerically the possibility of potentiostatic start up of a single cell to achieve

self-start from as low as -30°C in around 50 s under realistic conditions. Last but not least, they

concluded that though it may not be possible to apply this start up method directly to fuel cell, it

may be beneficial in the application of current-ramping strategies.Schießwohl et al. [16] explored

some significant parameters influencing the cold start ability of PEM fuel cell system in particular

preventing the formation of ice in and around catalyst layer. Successful cold start ability was

indicated and compared by the time taken for fuel cell system to achieve 50% of its maximum

power. Parameters involved were thermal mass, duration of shut down strategy, cell voltage and

start up temperature. The shutdown strategy was found to be a major parameter in cold start

success. As the cell dryness increases, the water absorbing capacity of the membrane also increases

which in turn prolong the time before membrane saturation. No performance loss occurs if the

saturation happens after the MEA temperature rise to at least 0°C.

In related research, Chacko et al. [17] investigated the high-frequency resistance behavior, water

motion, and ice accumulation in a PEM fuel cell before, during and after -10°C constant current

cold start experiments. Optimization in cold start performance occurred when the cell resistance

increased before start up which indicated PEM dehydration. During cold start, the cell resistance

increased due to PEM hydration and further increased due to ice formation in and around cathode

catalyst layer. At low current densities, super-cooled water was observed. They concluded that at

lower current densities, the PEM water storage capacity remained but the amount of ice formation

in and around cathode catalyst layer increased. At lower current densities, the amount of heat

generated was more, but the rate of heat generation was lower, than that of higher current densities.

Thus, there is an acceptable current density range that balances the quantity of heat generation and

the time needed to achieve successful cold start. In summary, researchers in general agree that the

following factors play significant roles in cold start ability of PEM fuel cells: (a) low membrane

water content is beneficial since dryer membrane will has more water-storage capacity and will be

able to prolong the duration of survival before freeze out, (b) dry shut down purge technique should

be applied since this will lead to dryer membrane, (c) an optimal range of startup current density

exists for each stack design and configuration, and (d) high frequency response of the cell decreases

during cold start and gradually increases at the end of cold start.

Conclusion

Extensive research had been conducted using experimental studies to fully understand the

behaviors and performance of fuel cell at sub-freezing temperatures. Nevertheless, there are still

significant issues and obstacles that need to be resolve before fully commercialize system can be

achieved. In all of the previous experimental researches [3], little consideration was given to the

nature of heat loss to the surroundings during sub-freezing start up. Some of the previous

researchers assumed that all heat generated in the fuel cell is used to warm up the product water by

thermally insulating the fuel cell; hence there is no heat loss to the environment. Others [11, 12]

acknowledged that there is heat loss to the surroundings but did not look into it in much detail. Most

of the experimental works [6, 15-16] done so far did not consider the increase of cell temperature

during start up. This may be due to the difficulties in obtaining accurate temperature measurements

of the fuel cell components. Based on the review, majority of the experimental work on PEM fuel

cell cold start were on single cell, while only a handful of researchers such as Oszcipok and

Bégot[6, 14] conducted sub-freezing PEM fuel cell stack start up. While successful single PEM fuel

cell start up was achieved from temperature as low as -30°C, it was not the case for PEM fuel cell

stack. Oszcipok found that cold start of a 30 W stack from -20°C was possible only with aid of

external energy. On the other hand, though Bégot managed to successfully self start up a 2 kW

stack from temperature of -5°C, the time taken was unacceptably long and he failed to start up the

stack at lower temperatures. Based on the current state of research involving stack cold start, further

research is necessary to fully understand the operation and mechanism of PEM fuel cell cold start.

854 Mechanical & Manufacturing Engineering

References

[1] Larminie J and Dicks A. Fuel Cell System Explained. John Wiley & Son, Inc. 2003

[2] U.S. Department of Energy. Progress Report for Hydrogen, Fuel Cells, and Infrastructure

Technologies Program (Appendix A). 2002.

[3] Schmittinger W and Vahidi A. A Review of the Main Parameters Influencing Long-Term

Performance and Durability of PEM Fuel Cells. Journal of Power Sources 180(1) (2008) 1-14.

[4] Siegel C. Review of Computational Heat and Mass Transfer Modeling in Polymer-

Electrolyte-Membrane (PEM) Fuel Cells. Energy 33(9) (2008) 1331-1352.

[5] Hishinuma Y, Chikahisa T, Kagami F. The Design and Performance of a PEFC at a

Temperature below Freezing. JSME International Journal Series B 47(2) (2004) 235-241.

[6] Oszcipok M, Zedda M, Riemann D. Low Temperature Operation and Influence Parameters on

the Cold Start Ability of Portable PEMFCs, Journal of Power Sources 154(2) (2006) 404-411.

[7] Oszcipok M, Zedda M, Hesselmann J. Portable Proton Exchange Membrane Fuel-Cell

Systems for Outdoor Applications. Journal of Power Sources 157(2) (2006) 666-673.

[8] Hottinen T, Himanen O, Lund P. Performance of Planar Free-Breathing PEMFC at

Temperatures Below Freezing. Journal of Power Sources 154(1) (2006) 86-94.

[9] Yan Q, Toghiani H, Lee Y. Effect of Sub-Freezing Temperatures on a PEM Fuel Cell

Performance, Startup and Fuel Cell Components. Journal of Power Sources 160(2) (2006)

1242-1250.

[10] Tajiri K, Tabuchi Y, Wang C. Isothermal Cold Start of Polymer Electrolyte Fuel Cells.

J.Electrochem.Soc. 154(2) (2007) B147-B152.

[11] Tajiri K, Tabuchi Y, Kagami F, Effects of Operating and Design Parameters on PEFC Cold

Start. Journal of Power Sources 165(1) (2007) 279-286.

[12] Pinton E, Fourneron Y, Rosini S. Experimental and Theoretical Investigations on a Proton

Exchange Membrane Fuel Cell Starting Up at Subzero Temperatures. Journal of Power

Sources 186(1) (2009) 80-88.

[13] Bégot S, Harel F, Kauffmann JM, Design and Validation of a 2 kW-Fuel Cell Test Bench for

Subfreezing Studies. Fuel Cells 8(2) (2008) 138-150.

[14] Bégot S, Harel F, Kauffmann JM. Experimental Studies on the Influence of Operational

Parameters on the Cold Start of a 2 kW Fuel Cell. Fuel Cells 8(2) (2008) 138-150.

[15] Jiang F and Wang C. Potentiostatic Start-Up of PEMFCs from Subzero Temperatures. Journal

of the Electrochemical Society 155(7) (2008) B743-B751.

[16] Schießwohl E, von Unwerth T, Seyfried F. Experimental Investigation of Parameters

Influencing the Freeze Start Ability of a Fuel Cell System. Journal of Power Sources 193(1)

(2009)107-115.

[17] Chacko C, Ramasamy R, Kim S. Characteristic Behavior of Polymer Electrolyte Fuel Cell

Resistance during Cold Start. Journal of the Electrochemical Society 155(11) (2008) B1145-

B1154.

Applied Mechanics and Materials Vol. 315 855

Mechanical & Manufacturing Engineering 10.4028/www.scientific.net/AMM.315 A Review of Experiments on Cold Start of PEM Fuel Cells 10.4028/www.scientific.net/AMM.315.851