ECS Transactions, 58 (1) 1601-1611 (2013)
10.1149/05801.1601ecst ©The Electrochemical Society
Pressure Drop and Voltage Response of PEMFC Operation under Transient
Temperature and Loading Conditions
Rupak Banerjee; Evan See; Satish G. Kandlikar
Department of Mechanical Engineering, Rochester Institute of Technology
76 Lomb Memorial Drive, Rochester, New York – 14623
Transportation powertrains need to be versatile in power
generation to meet the varying vehicular dynamic loads. The PEM
fuel cell stacks within vehicles are therefore designed to meet this
additional requirement under part-load and varying-load driving
conditions. The size of the on-board energy storage systems can be
reduced significantly if the fuel cell powertrain is able to generate
the load faithfully in corresponding to the varying demand load.
The voltage and pressure drop variations in response to changes in
the temperature and load conditions are experimentally studied in
the current investigation for characterizing the dynamic system
behavior. Temperature of the cell was changed over 20°C - 80°C
in short time periods (<600 seconds) and the load conditions were
changed by 400 mA/cm2 in 300 seconds. It was observed that
changes in load at lower temperatures have a significant impact on
the pressure drop in the channels. Additionally, a rapid increase or
decreases in temperature have an adverse effect on the cell
performance. Rapid increase in cell temperature results in
dehydration of the membrane, thereby reducing the performance;
while rapidly decreasing the cell temperature may result in
increased condensation.
Introduction
Proton Exchange Membrane Fuel Cells (PEMFCs) are seen as a viable replacement for
IC engines for automotive transportation. The PEM fuel cell stack within vehicles often
operates under part load and varying load conditions. Load conditions change based on
the driving, and can be very dynamic in nature. Designing the PEMFC powertrain to
follow the load faithfully would reduce the requirement for large energy storage systems
on board the vehicle (1, 2).
Uzunoglu and Alam (2) presented a solution to replace the large battery systems by
designing a control system with an ultra-capacitor in conjunction with the fuel cell to
meet the surge in power demand during acceleration of the vehicle. If the difference
between power generation and power demand is small, the ultra-capacitors would be able
to meet the power requirements. The ability of the fuel cell to closely match the power
demand will thus reduce the power requirement from the capacitor during transients.
Most studies reported in literature investigate and quantify the steady-state behavior of
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1601subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
fuel cells. Although the steady state behavior is important, automotive conditions are
transient in nature and therefore a clear understanding of the transient response of PEM
fuel cells needs to be developed. Very few experimental studies have been reported on
the performance characterization under transient conditions. In this regard, pressure drop
offers a good diagnostic tool for the two-phase flow in PEMFCs (3) with distinct pressure
drop signatures for each flow regime (4, 5), and can be utilized for characterizing the
transient performance.
Hamelin et al. (6) used a Ballard fuel cell stack, and showed that the response time of
load changes on the output voltage is less than one second. The theoretical time for the
electrochemical cell to respond is about 10 µs. He et al. (7) used pressure drop to predict
the flooding of electrodes and the porous media. They used interdigitated electrode
configurations in their study. Lu et al. (4) also showed that the pressure drop could be
used to identify flow patterns in the flow channels. Pressure drop is thus seen to be a
good diagnostic tool in characterizing the PEM fuel cells.
Wang and Wang (8) numerically investigated the transient effect of change in inlet
relative humidity on the water transport in the channels, GDLs and the membrane. They
varied the inlet humidification of both anode and cathode gas streams, and observed
water saturation in the GDLs and the membrane. The drying effect was studied; a lag of
about 20 seconds was reported for the membrane to dry. The anode side dried faster due
to the high diffusivity of water in the anode GDL.
Rabbani and Rokni (9) used a commercial fuel cell stack by Ballard Power Systems to
investigate the effect of transients. They varied the load on the stack from 60A to 100A,
effectively changing the current density from 0.21 A/cm2 to 0.35 A/cm2 in the process.
They observed that the voltage response was very rapid, matching with their original
hypothesis. However, the temperature increased due to the increased heat generation,
which resulted in transient temperature profiles. The temperature fluctuated between
60°C and 70°C, and did not reach steady state even after 5 minutes. An increase in the
load results in an increased heat generation within the cell. The coolant system is
supposed to regulate the temperature, but the lag in response leads to this transient
behavior. In a fuel cell vehicle, changing load conditions would affect the temperature of
the stack, and may lead to a lag in the power delivery.
A fuel cell starts at the ambient temperature, which varies from sub-freezing to midthirties in most regions. However, the optimal operating temperature for PEMFCs is
between 60°C and 80°C (10 – 14). The stack needs to reach this optimal operating
temperature range at the earliest to improve the efficiency as well as to meet the demand.
Auxiliary heaters are used along with the coolant circulation to raise the temperature
rapidly. In some cases, this causes a rapid change in temperature, which may also have an
adverse effect on the cell performance. Rabbani and Rokni (9) noted that the heat
generated in the cells also changes with the power output, and there is a time lag before
the coolant is able to bring the temperature back to the desired steady state value.
The gases entering the channels are partially humidified, while they are mostly saturated
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1602subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
towards the exit. Any rapid change in the temperature of the cell may result in
unexpected condensation of liquid water in the gas channels resulting in an increased
two-phase pressure drop, and a reduced performance of the cell. Also, the response from
the stored water in the GDL is dependent on the local conditions before and during the
transients. In the present work, the different conditions and cases discussed above are
investigated. Each test is designed to simulate a specific case discussed above. Different
load and temperature transients are imposed on the cell, and the response in terms of
voltage as well as two-phase pressure drop in the channels is observed.
Experimental Setup
An in-situ setup is used for the current investigation into the transient effects of varying
temperature and load conditions. A Greenlight Technology G40 fuel cell test stand is
used to control the flow of reactants and the load conditions to the cell. The test stand is
capable of controlling the flow rates, humidity, and gas temperature and is operated in
stoichiometry mode. This ensures a constant stoichiometry to be maintained as the air
flow is modulated according to the load being applied. The reactant supply line and dew
point temperature are controlled independently for the hydrogen and air lines. Figure 1
shows the test setup with the different components outlined.
Figure 1: In situ fuel cell setup showing the test stand, fuel cell, and the heated gas lines.
The fuel cell is built and assembled in-house, and has an active area of 50 cm2. A Gore
Membrane is used with MRC (Mitsubishi Rayon Corporation) 105 GDL on both sides.
The GDLs have an MPL coating on one side. Although wavy channels have been used in
the earlier steady state tests, Straight reactant channels are used in the present study to
deliver the reactants to the cell as described in Owejan et al. (15). The gas lines are
heated with separate controls to ensure that the gases arrive at the cell inlet at the
specified test conditions and avoid condensation of water from the humidifier in the gas
lines. Automated testing using the HyAl Fuel Cell Automated Testing System (FCATS)
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1603subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
has been used employed to implement ramped changes in load conditions over different
time segments.
Coolant loops are integrated into the compression plates on both the anode and cathode
sides, allowing for control of the cell temperature. Eight 0.8 kW powered in-line heaters
are used to control the coolant temperature and heat the cell with coolant loops. A 2 kW
heat exchanger cooling system is placed in-line for rapid cooling of the cell at a fast rate,
along with a temperature controlled water bath with 0.5 kW of additional cooling. These
controls are used to change provide the ramp changes in the cell temperature for the
investigation of transient temperatures.
Figure 2: In situ fuel cell setup showing the thermocouples and differential pressure
gauges.
Two-phase pressure drops in the reactant channels are measured with differential
pressure sensors from Honeywell®. The sensors have a range of 0 – 34 kPa with an
accuracy of 0.25% of the range, resulting in an uncertainty of ±0.08 kPa. The cell also
incorporates individual channel pressure drop measurement in the entrance region for
measuring the instantaneous gas flow rate through each channel using the Hornbeck
equation, as described in Kandlikar et al. (15). However, the individual channel flow
rates are not presented in the current work.
Two arrays of five thermocouples each on the anode and cathode sides (in the
compression plates) monitor the temperature of the cell. The coolant lines have
thermocouples at the end of the in-line heaters to regulate the power supplied to the inline heaters for controlling the cell temperature. The gas lines have separate
thermocouples to control the temperature of the incoming reactants. Figure 2 shows the
cell with the control thermocouples. The thermocouples have been calibrated to read
within an uncertainty of less than ±1°C.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1604subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
Load and cell voltages are recorded by the G40 test stand at a rate of 1 Hz. Temperature
of the cell and the pressure drop in the channels on the cathode and anode sides of the cell
are recorded on an auxiliary computer through a DAQ comprising of a PXIe series
chassis from National Instruments.
Test Matrix and Test Protocol
In order to improve the repeatability of the test results, a standard operating procedure
was utilized with a 2.5 hour cell conditioning routine, performed at the start of every
testing day. This ensured the membrane was adequately hydrated, and the cell was at a
steady state condition prior to the testing.
For transient load testing, the cell was operated at steady state before changing the load
through a ramp up or ramp down procedure. Different load changes were implemented
over a 300 second period using the HyAl FCATS automation system. The voltage
response and the two-phase pressure drop response were recorded. Tables 1 and 2 show
the different test conditions investigated in this study.
Table 1: Tests for transient effect from changing temperatures
Temperature Start
Temperature End
(°C)
(°C)
40
60
40
80
60
40
80
40
Table 2: Tests for transient effect from changing load
Load Start
Load End
2
(A/cm )
(A/cm2)
1.0
0.6
0.6
1.0
Current Densities
(A/cm2)
1.0
1.0
1.0
1.0
Temperature of Test
(°C)
40, 60, 80
40, 60, 80
In addition to the above mentioned tests, additional experiments were performed to check
the repeatability of the results. Long duration tests were run to understand the time scale
over which the transients play a role. These are discussed in the results section.
Results
Figure 3 shows the cathode channel pressure drop over a period of about 3 hours. After
operating the cell at 0.8 A/cm2 for the first 2 hours, the load condition is changed to 0.1
A/cm2 to investigate the time it requires for the system to return to steady state in terms
of the two-phase pressure drop in the channels. Although the pressure has been allowed
to stabilize for over 6000 seconds (1.6 hours), there are still some inherent fluctuations in
the pressure drop signature (Region A). As soon as the load is changed, from 0.8 A/cm2
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1605subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
to 0.1 A/cm2, the flow rate of reactants is reduced, and the pressure drop decreases in the
channels. Although there is an initial dip in the pressure drop (Region B), it slowly starts
to increase, and reaches a new steady state value. The time required for the pressure drop
to regain its steady state behavior is the time duration over which the transient behavior is
relevant.
Figure 3: Time required for the two-phase pressure drop to reach steady state after the
system undergoes a change in load. The load was reduced from 40A to 5A.
Next, the results from transient load conditions are discussed. The load changes have
been implemented using a ramped profile over a period of 300 seconds. Figure 4 shows
three different cases where the current density is reduced from 1.0 A/cm2 to 0.6 A/cm2.
The response is noted from the voltage behavior and the two-phase pressure drop
signature. The three different figures compare the results from the tests conducted at
temperatures of 40, 60 and 80°C. In each of the tests, the stoichiometry is held constant at
1.5 for anode and 2.0 for cathode. It can be seen that the voltage response and the twophase pressure drop responses are almost instantaneous. This is expected, and is in line
with the results of Hamelin et al. (6). The pressure drop decreases as consumption and air
flow rate decrease, while the voltage increases with a decrease in the ohmic resistance
due to lower current being drawn. It can also be seen that the pressure drop response with
time is much steeper at higher temperatures, as the final pressure drop remains higher at
the lower temperatures. There is significant water in the channels which results in the
higher pressure drop. However, at the higher temperatures, most of the water is
transported in the vapor phase and therefore the pressure drop is lower. Thus the pressure
drop slope is much steeper at the higher temperatures.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1606subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
Figure 4: Effect of decreasing load on the two-phase pressure drop and the voltage
response. The load was reduced from 50 A to 30 A for each case shown. a) Response at
40°C, b) response at 60°C and c) response at 80°C.
Figure 5 shows the effect of increasing load on the voltage response and the two-phase
pressure drop in the reactant channels, as the load is increased from 0.6 A/cm2 to 1.0
A/cm2. The plots show the response at three different temperatures of 40, 60 and 80°C.
Similar to the observations for decreasing current density (Figure 4), the results show
here as well that the pressure drop slopes are steeper when the load transition takes place
at the higher cell temperatures. Additionally, it may be noted that at 40°C, there is a
distinct peak at the end of the transition after which the two-phase pressure drop
decreases and approaches the steady state value. This is similar to the trough observed in
Figure 3. However, with a significantly lower amount of liquid present in the channels,
the same peak is no longer observed at the higher temperatures. Therefore, transient load
changes have a greater impact on two-phase flow at the lower temperatures as compared
to the higher temperatures.
Next, the impact of a change in temperature is studied on the fuel cell performance. The
optimum PEM operation is in the range of 60 – 80°C. A higher temperature results in
reduced activation losses, and therefore improves the voltage response from the cell.
Thus, it would be beneficial to operate the fuel cell at higher temperatures. However, at
startup, the temperature of the cell is at the ambient conditions. Therefore, at the start of
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1607subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
operation for a PEMFC powered vehicle, the cell would need to transition from the
ambient temperature to the optimum temperature of 80°C. The following discussion
brings forth the issues that may arise, and the transient behavior that needs to be
understood during the transition of cell temperature from ambient to 60 or 80°C.
Figure 5: Effect of increasing load on the two-phase pressure drop and the voltage
response. The load was increased from 30 A to 50 A for each case shown. a) Response at
40°C, b) response at 60°C and c) response at 80°C.
Figure 6 shows the effect of dynamically increasing temperature on the cell performance,
and the two-phase pressure drop in the cathode reactant channels. Figure 6(a) shows the
response for a temperature change from 40 to 60°C, which is accompanied by an increase
in voltage response for the steady load. The pressure drop in the channels decreases as
more of the water is removed in the vapor phase, and therefore liquid blockages are
reduced. Figure 6(b) shows the response for a temperature change from 40 to 80°C. The
voltage reaches a peak just before 60°C, and then starts to decrease. The pressure drop
also decreases at first, and then begins to increase. This reversal, as the cell temperature
continues to increase beyond 60°C, may be attributed to the loss in hydration of the
membrane. The rapidly increasing cell temperature results in an increased saturation
temperature. Therefore most of the water is removed from the cell. The humidification
systems respond to the change in temperature at a slower rate, and therefore cannot
provide the necessary humidification required for the cell to operate at 80°C. The
increase in pressure drop is due to the increased gas velocities, resulting from a lower gas
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1608subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
density at 80°C. There is no more liquid water in the channels, and therefore the increase
in saturation temperature has a less significant effect on the two-phase pressure drop.
Figure 6: Effect of increasing temperature on the voltage response and the two-phase
pressure drop in the channels. The cell is operated at a constant load of 50A with a
stoichiometry of 1.5 on the anode side and 2 on the cathode side. a) cell temperature
increases from 40 to 60°C b) cell temperature increases from 40 to 80°C.
Figure 7: Effect of decreasing temperature on the voltage response and the two-phase
pressure drop in the channels. a) Cell temperature decreases from 80 to 40°C and b) cell
temperature decreases from 75 to 55°C.
Figure 7 shows the effect of dynamically decreasing temperature on the cell performance,
and the two-phase pressure drop in the cathode reactant channels. Figure 7(a) shows the
decrease in temperature from 77°C to 40°C, while figure 7(b) shows the decrease in
temperature from 75°C to 55°C. It can be observed from these figures that the decrease in
temperature from 77°C to 40°C causes a significant decrease in voltage, and thus the
performance suffers due to the rapid decrease in temperature. The pressure drop however
does not drop as quickly. This can be attributed to condensation in the channels for the
humidified inlet conditions. Additionally, the cell is cooled at the compression plates,
while the MEA remains hotter due to the reactions resulting in temperature gradient
driven water transport into the channels. These two factors result in the pressure drop
remaining high in the channels, even though the temperature has decreased. However, for
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1609subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
a temperature transition from 75°C to 55°C, i.e. a drop of 20°C, the effect is not as
significant. There is a negligible drop in the voltage response and a very small reduction
in the pressure drop in this case.
Summary
The current investigation explores the impact of two-phase flow on the transient
performance of PEM fuel cells. Previous works have shown that the electrochemical
reaction has a transient behavior lasting less than 1 second. However, transient response
in two-phase flow lasts for time periods on the order of several minutes. Therefore, the
effect of the transient behavior of two-phase flow on the performance of a PEM fuel cell
is investigated here.
The key findings are summarized as follows:
 In response to a change in operating conditions in a PEMC, time durations on the
order of several minutes are required for the two-phase flow in the reactant
channels to return to steady state behavior.
 Transient load changes have a greater impact on two-phase flow at the lower
temperatures.
 Decreasing temperature rapidly below 60°C results in a loss of performance from
the cell, and increased pressure drop due to increased liquid water from
condensation within the cell.
 Increasing temperature rapidly beyond 60°C results in a loss of performance that
could be attributed to the membrane dryout due to inability of the humidification
system to respond quickly as the cell temperature increases.
 Temperature fluctuations between 60 and 80°C are not significant, as the water
remains mostly in vapor phase in both cases.
The effect of two-phase pressure drop on the transient performance of a PEM fuel cell
needs to be explored further. A detailed investigation is required to establish the time
constant for the transient response, and the parameters affecting the time constant.
Acknowledgments
This work was conducted in the Thermal Analysis, Microfluidics and Fuel Cell
Laboratory at the Rochester Institute of Technology. The work was supported by the US
Department of Energy under the award number DE-EE0000470. The authors would like
to thank Wenbin Gu and Jeffrey Gagliardo from the Electrochemical Energy Research
Laboratory at General Motors for supplying the GDL samples tested in this work and for
general technical discussions facilitating this work.
References
[1] P. Corbo, F. E. Corcione, F. Migliardini, and O. Veneri, “Experimental study of a
fuel cell power train for road transport application,” J. of Power Sources, vol. 145,
no. 2, pp. 610–619, Aug. 2005.
[2] M. Uzunoglu and M. S. Alam, “Dynamic modeling, design, and simulation of a
combined PEM fuel cell and ultracapacitor system for stand-alone residential
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1610subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 1601-1611 (2013)
applications,” Energy Conversion, IEEE Transactions on, vol. 21, no. 3, pp. 767–
775, Sep. 2006.
[3] J. Wu, X. Zi Yuan, H. Wang, M. Blanco, J. J. Martin, and J. Zhang, “Diagnostic tools
in PEM fuel cell research: Part II: Physical/chemical methods,” Int. J. of Hydrogen
Energy, vol. 33, no. 6, pp. 1747–1757, Mar. 2008.
[4] Z. Lu, S. G. Kandlikar, C. Rath, M. Grimm, W. Domigan, A. D. White, M.
Hardbarger, J. P. Owejan, and T. A. Trabold, “Water management studies in PEM
fuel cells, Part II: Ex situ investigation of flow maldistribution, pressure drop and
two-phase flow pattern in gas channels,” Int. J. of Hydrogen Energy, vol. 34, no.
Compendex, pp. 3445–3456, 2009.
[5] M. Grimm, E. J. See, and S. G. Kandlikar, “Modeling gas flow in PEMFC channels:
Part I – Flow pattern transitions and pressure drop in a simulated ex situ channel with
uniform water injection through the GDL,” Int. J. of Hydrogen Energy, vol. 37, no.
17, pp. 12489–12503, Sep. 2012.
[6] J. Hamelin, K. Agbossou, A. Laperrière, F. Laurencelle, and T. . Bose, “Dynamic
behavior of a PEM fuel cell stack for stationary applications,” Int. J. of Hydrogen
Energy, vol. 26, no. 6, pp. 625–629, Jun. 2001.
[7] W. He, G. Lin, and T. Van Nguyen, “Diagnostic tool to detect electrode flooding in
proton-exchange-membrane fuel cells,” Aiche J., vol. 49, no. 12, pp. 3221–3228,
2003.
[8] Y. Wang and C.-Y. Wang, “Two-Phase Transients of Polymer Electrolyte Fuel
Cells,” J. of The Electrochem Soc, vol. 154, no. 7, p. B636, 2007.
[9] R. A. Rabbani and M. Rokni, “Dynamic characteristics of an automotive fuel cell
system for transitory load changes,” Sustainable Energy Technologies and
Assessments, vol. 1, no. 1, pp. 34–43, 2013.
[10] G. Zhang and S. G. Kandlikar, “A critical review of cooling techniques in proton
exchange membrane fuel cell stacks,” Int. J. of Hydrogen Energy, vol. 37, no. 3, pp.
2412–2429, Feb. 2012.
[11] S. G. Kandlikar and Z. Lu, “Thermal management issues in a PEMFC stack–A
brief review of current status,” App. Thermal Engineering, vol. 29, no. 7, pp. 1276–
1280, 2009.
[12] S. Yu and D. Jung, “Thermal management strategy for a proton exchange
membrane fuel cell system with a large active cell area,” Renewable Energy, vol. 33,
no. 12, pp. 2540–2548, Dec. 2008.
[13] E. Endoh, S. Terazono, H. Widjaja, and Y. Takimoto, “Degradation Study of
MEA for PEMFCs under Low Humidity Conditions,” Electrochem. Solid-State Lett.,
vol. 7, no. 7, p. A209–A211, Jul. 2004.
[14] C. Spiegel, Designing and Building Fuel Cells. Mcgraw-hill, 2007.
[15] J. P. Owejan, J. J. Gagliardo, J. M. Sergi, S. G. Kandlikar, and T. A. Trabold,
“Water management studies in PEM fuel cells, Part I: Fuel cell design and in situ
water distributions,” Int. J. of Hydrogen Energy, vol. 34, no. 8, pp. 3436–3444, May
2009.
[16] S. G. Kandlikar, Z. Lu, W. E. Domigan, A. D. White, and M. W. Benedict,
“Measurement of flow maldistribution in parallel channels and its application to exsitu and in-situ experiments in PEMFC water management studies,” Int. J. of Heat
and Mass Transfer , vol. 52, no. 7–8, pp. 1741–1752, Mar. 2009.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
1611subject to ECS license or copyright; see ecsdl.org/site/terms_use