Lifetime Evaluating and the Effects of Operation Conditions on Automotive Fuel Cells

PEI Pucheng, YUAN Xing, LI Pengcheng, , CHAO Pengxiang, and CHANG Qianfei

1 State Key Laboratory of Automotive Safety and Energy, Tsinghua University, Beijing 100084, China

2 School of Transportation Science and Engineering, Beihang University, Beijing 100191, China

1 Introduction

Proton exchange membrane fuel cell(PEMFC) has many advantages compared with traditional internal combustion engine(ICE) in making alternative power system for vehicles. GRANOVSKII, et al[1], studied life cycle of hydrogen fuel cell and gasoline vehicles. WANG, et al[2],found that fuel-cell vehicles could achieve the envisioned energy and emission reduction benefits by carefully examined pathways for producing the fuels. THOMAS[3]and COLELLA, et al[4], also found fuel cell emission and energy advantage. Nevertheless, the durability of PEMFC is much lower in automotive application than in stationary application[5–7]which has been becoming the obstacle to the fuel cell development.

Normally, it requires big expenditure and much time to evaluate fuel cell lifetime in ordinary way. There have also been reports of a 26 300 h single cell life test operated with a membrane electrode assembly(MEA) in stationary fuel cell applications. The performance degradation rate of the cell was determined between 4 μV/h and 6 μV/h at the operating current density of 800 mA/cm2, which costs 3 years just to get these results[8]. Certainly, the more rigorous the operation is, the fewer test hours are taken.However, it is far away from the fuel cell real working conditions so that the result reliability used to evaluate fuel cell lifetime needs to be analysed seriously[9–11].

Study on lifetime compared with different fuel cells is significantly in favor of finding out the mechanism of fuel cell degradation. And two PEM fuel cells’ durabilities in different operation[5]were compared and investigated. It needs to be concerned how the degradation happened under the same operation condition.

It is confirmed that the degradation of fuel cell vary with different operation condition, for instance, fuel cell working load, fuel cell idle load, etc[5,12–13]. It implies that optimizing the fuel cell working load makes it possible to gain the external lifetime.

In this paper, automotive fuel cell driving cycles was ascertained based on the real loading map of a fuel cell bus in urban road test. A proposal of vehicular fuel cell lifetime evaluating method was given, and two PEMFC stacks were tested and their lifetimes were evaluated in laboratory.Thereby, all operating condition contributors to fuel cell lifetime degradation were gained to help to optimize the operation mode.

2 Lifetime Quick Evaluating Method

2.1 Definition about the end of automotive fuel cell lifetime

The average fuel cell voltage is often 0.7 V at rated condition. We define that the lifetime of this automotive fuel cell is terminated when the cell voltage decreases 0.07 V or 10% from the start rated point at the same current[14]. Fig. 1 shows the lifetime start I-V curve to the end I-V curve test on a fuel cell stack.

Fig. 1. Fuel cell lifetime defined on a real fuel cell bus

2.2 Equation of fuel cell lifetime evaluating

Running on a fixed route every day, one of our demonstrating fuel cell buses has covered 43 000 km range.With considering the loading map of the bus, a laboratory test driving cycle simulating vehicle driving cycle is drawn out as following Fig. 2 and Table 1, including 13 min high power condition, 14 min idle condition, 56 load changing cycles and one time start-stop in 1 h.

Fig. 2. Laboratory test cycle simulating driving cycle

Table 1. Fuel cell working status of the laboratory test

The degradation of automotive fuel cell is complex,however, it is dedicated to above four working status mentioned before. It’s known the degradation rate of fuel cell performance is linear, and the equation of fuel cell lifetime can be calculated in the following expressions[14]:

Where P1,′ P2′, P3,′ 4P′are performance degradation rates resulted in by load change cycles, idle condition, high power load condition and start-stop cycles, respectively,measured in laboratory, and the means of n1, n2, t1, t2are shown in Table 1. ?P is the maximal allowed degradation of voltage which is 0.07 V. k is the accelerating coefficient which due to the difference between laboratory and road. In Ref. [14] it is 1.72, but the calculated lifetime shows 10%less than the road test lifetime. So it is taken as 1.6 in this paper.

In four laboratory tests, namely, load change cycles test,start-stop cycles test, idle condition test and high power load condition test, the fuel cell lifetime can be calculated by Eq. (1).

3 Quick Lifetime Evaluating on Two Fuel Cell Stacks

3.1 Experiment of two fuel cell stacks

Two different fuel cell stacks which have different flow field but the same active area are evaluated by the lifetime quick evaluating method. No. 1 stack and the fuel cells of demonstrating bus are completely identical. First of all, the two stacks are tested by the laboratory driving cycles. And then, the lifetimes of the two stacks are both calculated as shown in Fig. 3 in which Fig. 3(a) presents the No. 1 fuel cell stack lifetime degradation by laboratory driving cycles.And Fig. 3(b) shows the No. 2 stack’s test result.

Fig. 3. Two stack laboratory driving cycle tests

The degradation of laboratory driving cycles test can be calculated by Fig. 3:

So, the lifetime of the two stacks running in the former driving cycles can be gained directly considering the definition of the end of automotive fuel cell lifetime: Lfc1=1 080 h, Lfc2=750 h.

The contribution to fuel cell voltage degradation by load changing cycles is presented in Fig. 4. Stack current changes from 23 A to 98 A and then to 23 A repeatedly while the load changing cycle test. The voltage decay rates can be measured from Fig. 4 as follows:

Fig. 4. Voltage degradation by load changing cycles

Fig. 5(a) shows 50 h test result of No. 1 stack, in which the idling current density is 10 mA/cm2and the fuel cell performance gets almost full recovery at every beginning,with a little decay rate beyond retrieve. It is significative that although we took test in irregular way for 10 h after 25 h, the following test results show the same changing rate as the former test. To enhance the decay rate accuracy, it is important to keep test process regularly and strictly.Fig. 5(b) presents the No. 2 stack test result, which the idling current density is 10 mA/cm2and the experiment data is good to be accepted. From these figures, we get the voltage decay rates as follows:

Fig. 5. Voltage degradation by idling cycles

The high power cycles also affect the fuel cell lifetime,shown in Figs. 6(a) and 6(b).

Fig. 6. Voltage degradation by high power cycles

The two fuel cell stacks both work at current of 100 A in the test status, and then the polarization curve is measured.

The decay rates are as follows:

Fig. 7 presents the degradation caused by start-stop operation in No. 1 fuel cell stacks. After a few times of start-stop operation, the stacks voltages are tested at current 100 A as same as load changing test. The degradation values can be gained from Fig. 7:

Fig. 7. Voltage degradation by start-stop cycles in No. 1 stack

It is noted that the No. 2 stack’s performance shows nonlinear decay, because of the water pump in the test platform stopped several times in unknown reason.

We found the phenomena in No. 1 stack as

So we can use Eq. (12) to get the degradation value of No. 2 stack caused by start-stop cycles:

Those mean that we can achieve the fuel cell lifetime just by the four tests of driving cycles, load changing cycles,idling cycles and high power cycles, and the total test time is no more than 250 h.

3.2 Lifetime calculating and analysis

The fuel cell voltage degradation rates of No. 1 fuel cell stack by load change cycles, idle condition, high power load condition and start-stop cycles separately were shown as Eqs. (4), (6), (8) and (10). Eqs. (5), (7), (9) and (13)show the voltage degradation rates of No. 2 fuel cell stack.

Fig. 8 shows the voltage decay rate difference in the two stacks. In the No. 1 stack, the load change cycling and the start-stop cycling are main factors contributing to fuel cell performance decay. One third of deterioration is resulted in by start-stop cycling and over 50% is by load change cycling. By modifying start-stop cycling and load change cycling or decreasing their times, the fuel cell lifetime will be prolonged undoubtedly. Table 2 shows the optimization of working conditions of No. 1 stack and the predicted lifetimes in fuel cell buses.

Fig. 8. Comparing of different operations between two stacks

Table 2. Optimization of working conditions

Fig. 9 shows the degradation rate of the No. 1 fuel cell bus tallies with the predicted voltage decay rate, further proving the validity of Eq. (1).

Fig. 9. No. 1 fuel cell bus predicted lifetime

4 Ascertainment about Best Running Load of Automotive Fuel Cell

The voltage decay rate of high power cycles tested at 70 A and measured at 100 A is shown in Fig. 10. Compared with Fig. 6(b), the voltage decay rate of high power cycles at 70 A which is 224 μV/h is higher than 110 μV/h at 100 A. This is likely due to the design of fuel cell flow field in which the current set of 100 A approaches the rated load so that the water and thermal management is better in all operation conditions.

Fig. 10. Voltage degradation by high power cycles tested at 70A

Fig. 11(a), Fig. 5(b), and Fig. 11(b) show the idle cycles test results at different currents density of 30 mA/cm2,10 mA/cm2and 5 mA/cm2. Results present that the lower the idle current is, the smaller the voltage decay rate is got.It is unexpected that the voltage grows up day by day tested by idle cycles at 1.4 A (5 mA/cm2). So we can use this character to prolong fuel cell lifetime.

Fig. 11. Voltage degradation by idle cycles test

Fig. 12 presents what the current set is chosen to make the lifetime of automotive fuel cell better. It implies that when the fuel cell works at idle condition, the lower load current is better for the fuel cell lifetime. And when it works at high power condition, the load current is around the rated set which is around the rated load to ensure the fuel cell has longer lifetime.

Fig. 12. Ascertainment about best running load

5 Conclusions

(1) The lifetime formula including of performance decay rates resulted by start-stop cycling, idling cycling, load change cycling and high power load cycling shows feasible as compared with the real urban road test of fuel cell bus.

(2) The automotive fuel cell lifetime can be gained based on Eq. (1) with no more than 250 h test in laboratory.

(3) The automotive fuel cell lifetime can be extended from 1 100 h to 2 600 h by optimizing operation conditions.

(4) Micro-current operation can prolong fuel cell lifetime.

[1] GRANOVSKII M, DINCER I, ROSEN M A. Life cycle assessment of hydrogen fuel cell and gasoline vehicles[J]. International Journal of Hydrogen Energy, 2006, 31(3): 337–352.

[2] WANG Michael. Fuel choices for fuel cell vehicles: well-to-wheels energy and emission impacts[J]. Journal of Power Sources, 2002,112(1): 307–321.

[3] THOMAS C E. Fuel cell and battery electric vehicles compared[J].International Journal of Hydrogen Energy, 2009, 34(15): 6 005–6 020.

[4] COLELLA W G, JACOBSON M Z, GOLDEN D M. Switching to a U.S. hydrogen fuel cell vehicle fleet: The resultant change in emissions, energy use, and greenhouse gases[J]. Journal of Power Sources, 2005, 150(4): 150–181.

[5] WAHDAME B, CANDUSSO D, FRANC-OIS X, et al. Comparison between two PEM fuel cell durability tests performed at constant current and under solicitations linked to transport mission profile[J].International Journal of Hydrogen Energy, 2007, 32(17): 4 523–4 536.

[6] ZHANG Shengsheng, YUAN Xiaozi, WANG Haijiang, et al. A review of accelerated stress tests of MEA durability in PEM fuel cells[J]. International Journal of Hydrogen Energy, 2009, 34(1):388–404.

[7] SCHMITTINGER Wolfgang, VAHIDI Ardalan. A review of the main parameters influencing long-term performance and durability of PEM fuel cells[J]. Journal of Power Sources, 2008, 180(1):1–14.

[8] CLEGHORN S J C, MAYFIELD D K, MOORE D A, et al. A polymer electrolyte fuel cell life test: 3 years of continuous operation[J]. Journal of Power Sources, 2006, 158(1): 446–454.

[9] WAHDAME Bouchra, CANDUSSO Denis, HAREL Fabien, et al.Analysis of a PEMFC durability test under low humidity conditions and stack behaviour modelling using experimental design techniques[J]. Journal of Power Sources, 2008, 182(2): 429–440.

[10] AKIRA Taniguchi, TOMOKI Akita, KAZUAKI Yasuda, et al.Analysis of degradation in PEMFC caused by cell reversal during air starvation[J]. International Journal of Hydrogen Energy, 2008,33(9): 2 323–2 329.

[11] FOWLER M, AMPHLETT J C, MANN R F, et al. Issues associated with voltage degradation in a PEMFC[J]. Journal of New Materials for Electrochemical Systems, 2002, 5(4): 255–262.

[12] LIN R, LI B, HOU Y P, et al. Investigation of dynamic driving cycle effect on performance degradation and micro-structure change of PEM fuel cell[J]. International Journal of Hydrogen Energy, 2009,34(5): 2 369–2 376.

[13] KULIKOVSKY A A, SCHARMANN H, WIPPERMANN K.Dynamics of fuel cell performance degradation[J]. Electrochemistry Communications, 2004, 6(1): 75–82.

[14] PEI Pucheng, CHANG Qianfei, TANG Tian. A quick evaluating method for automotive fuel cell lifetime[J]. International Journal of Hydrogen Energy, 2008, 33(14): 3 829–3 836.