CC BY
14Статья поступила в редакцию 12.08.2010. Ред. per. № 85Э
The article has entered in publishing office 12.08.2010. Ed. reg. No. 853
FCV DEVELOPMENT AT NISSAN K. Yoshizawa, R. Shimoi, K. Ikezoe, T. Aoyama, T. Arai, A. Iiyama
Nissan Motor CO., LTD. 1, Natsushima-cho, Yokosuka-shi, Kanagawa 237-8523, JAPAN +81-46-867-5348/+81-46-867-5332; y-koudai@mail.nissan.co.jp
With the aim of dramatically reducing CO2 emissions from automobiles, we are developing both electric vehicles (EVs) and fuel cell vehicles (FCVs) as zero emission vehicles. Research and development work on Nissan FCVs has been under way since 1996. The main issues that must be addressed to commercialize FCVs include: (1) to prevent performance decay, (2) to increase power density, (3) to improve subzero start-up capability, and (4) to develop technologies for reducing costs. This paper presents the results achieved to date and describes various activities under way to address these four main issues.
СОЗДАНИЕ АВТОМОБИЛЯ НА ТОПЛИВНЫХ ЭЛЕМЕНТАХ В КОНЦЕРНЕ «НИССАН» К. Йошизава, Р. Шимои, К. Икезое, Т. Аояма, Т. Араи, А. Иияма
В целях радикального снижения автомобильных выбросов CO2 нами ведется разработка как электромобилей, так и автомобилей на топливных элементах, которые характеризуются нулевым выбросом вредных веществ. Научные исследования и разработки в области создания автомобилей на топливных элементах в концерне «Ниссан» ведутся с 1996 г.. Основными задачами, которые необходимо решить для коммерческого освоения автомобилей на топливных элементах, являются:
1) предотвращение ухудшения рабочих характеристик;
2) увеличение плотности потока энергии;
3) усовершенствование системы пуска при отрицательных температурах;
4) разработка технологий для снижения стоимости.
В данной статье представлены полученные на сегодняшний день результаты, а также деятельность по решению этих четырех основных задач.
Organization(s): Manager, EV System Laboratory, Nissan Research Center, Nissan Motor CO., LTD., Ph.D. in Mechanical Engineering, Professional Engineer (Mechanical Engineering), 2 best paper awards of ASME, best presentation award of ASME, 2 best paper awards of JSAE, good patent award of Japan Invention Society.
Education: Hokkaido University (1985-1991).
Experience: Nissan Motor CO., LTD. Engineer (1991-2003), Nissan Motor CO., LTD. Manage(2004-2007), New Energy and Industrial Technology Development Organization (temporary assignment) Chief Engineer (2007-2010), Nissan Motor CO., LTD. Manager (2010).
Main range of scientific interests: fuel cells, batterys, internal combustion engines, flow and heat transportation phenomena, chemical reactions, etc.
Publications: ASME Transactions, JSME Transactions, SAE Transactions, JSAE Transactions, J. of ECS, etc.
Koudai Yoshizawa
A
Ryoichi Shimoi
Organization(s): Engineer, Advanced Materials Laboratory, Nissan Research Center, Nissan Motor CO., LTD.
Education: Tokyo Institute of Technology, Department of Mechanical Engineering and Science (1996-2000) Department of Mechanical and Control Engineering (2000-2002). Experience: Nissan Motor CO., LTD., Engineer (2002).
Main range of scientific interests: Fuel cells, heat transportation phenomena, Visualization, electrochemistry.
Publications: SAE Transaction.
Organization(s): Manager, EV System Laboratory, Nissan Research Center, Nissan Motor CO., LTD.
Education: Waseda University, Mechanical engineering (1989-1995).
Experience: Nissan Motor CO., LTD. Engineer (1995-2009). Nissan Motor CO., LTD., Manager (2009).
Main range of scientific interests: Fuel cells, Fuel Cell Power Plant System. Publications: SAE Transaction, JSAE Transaction.
Ikezoe-san
Takashi Aoyama
I
Takayuki Arai
Akihiro Iiyama
Organization(s): Manager, Research Testing Section No.1, Prototype and Test Department, Nissan Research Center, Nissan Motor CO., LTD.
Education: Waseda University, Mechanical engineering (1983-1989).
Experience: Nissan Motor CO., LTD. Engineer (1989-2004). Nissan Motor CO., LTD., Manager (2005).
Main range of scientific interests: Fuel cells, Internal combustion engines, Thermodynamics, Exhaust-gas aftertreatment system.
Publications: SAE Transaction, JSAE Transaction.
Organization(s): Senior Manager, EV System Laboratory, Nissan Research Center, Nissan Motor CO., LTD.
Education: Musashi Institute of Technology, Department of Mechanical Engineering (1982-1985).
Experience: Nissan Motor CO., LTD. Engineer (1985-2001), Nissan Motor CO., LTD., Manager (2002-2008), Nissan Motor CO., LTD., Senior Manager (2009).
Main range of scientific interests: Fuel cells, Internal combustion engines, Main moving parts design and test for Internal combustion engines.
Publications: JSME Transactions, JSAE Transactions, FISITA Transactions.
Organization: General Manager, EV System Laboratory, Nissan Research Center, Nissan Motor CO., LTD. Doctor of Engineering in Mechanical Engineering.
Education: 1982 Graduated from University of Tokyo, MS in Mechanical engineering.
Experience: 1982 joined Nissan Motor CO., LTD. Central Research Laboratory, Engaged in diesel engine and gasoline engine research, 1986-88, Visiting Industrial Fellow at the University of California at Berkeley (engine combustion), 1991 Doctor of Engineering (University of Tokyo, Mechanical Engineering), 2001 Joined Nissan FCV program and engaged in MEA and Stack development, 2008 General Manager of Fuel Cell Laboratory, 2010 General Manager of EV System Laboratory.
Main range of scientific interests: Internal Combustion engines, Spark ignition engines, Diesel engines, Fuel Cells.
Publications: SAE Transaction, JSAE, JSME.
Introduction
Rising levels of greenhouse gas emissions, especially carbon dioxide (CO2), are regarded as one of the causes of global warming in recent years. Activities by vehicle manufacturers to develop technologies for reducing CO2 emissions play an important role. In particular, programs are under way to develop electric vehicles (EVs) and fuel cell vehicles (FCVs), which are seen as being promising future powertrain systems because they operate on electricity or hydrogen as their motive power sources that can be produced from renewable forms of energy.
Both EVs and FCVs can play important roles as zero-emission vehicles that emit neither CO2 nor exhaust pollutants. In real-world driving, these vehicles can be used in ways that maximize the benefits of their respective advantages. As shown in Fig. 1, using EVs as small vehicles for short-distance trips and FCVs as large vehicles for traveling longer distances is thought to be effective in terms of their present technical capabilities [1]. At Nissan, we are planning to begin selling a mass-produced EV near the end of 2010. Development work on FCVs is proceeding toward a target market introduction in the early 2010s. The following sections describe the history and current status of FCV development at Nissan and issues that remain to be addressed.
Vehicle Size xee
.....
Large
/ FCV
Medium
Small
EV
Trip Distance
Short
Medium
Long
Fig. 1. Example of the positioning of EVs and FCVs based on a consideration of current technical capabilities [1]
History of FCV Development at Nissan
FCV Development from 1996 to 2005 Nissan initiated R&D work on FCVs in 1996 and built its first prototype vehicle equipped with a methanol reformer system in 1999. That was followed by the development of the XTERRA FCV in 2001, which used high-pressure hydrogen as its fuel source. The XTERRA FCV was supplied to the California Fuel Cell Partnership (CaFCP) program for use in conducting public-road driving tests. In 2002, a prototype FCV based on the X-TRAIL sport utility vehicle was developed and approved by the
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Minister of Land, Infrastructure and Transport (Fig. 2). Nissan then began public-road testing of the vehicle under the Japan Hydrogen & Fuel Cell Demonstration Project (JHFC).
Fig. 2. X-TRAIL FCV prototype [2]
A 2003 X-TRAIL FCV was then completed that was fitted with a newly developed fuel cell system, and limited leasing of the vehicle was launched in December 2003. Subsequently, a 2005 X-TRAIL FCV was newly developed (Fig. 3) that was approved by the Minister of Land, Infrastructure and Transport in December 2005 [2]. Nissan began leasing the vehicle to the Kanagawa prefectural government and to the city of Yokohama in April 2006, as the successor to the 2003 X-TRAIL FCV that had been leased to them on a limited basis. Testdrive sessions that enabled ordinary customers to actually test-drive an FCV were held every other week at Nissan's corporate headquarters in Tokyo. In addition, the X-TRAIL FCV was put in service in February 2007 as the world's first commercial FCV taxi , thereby providing opportunities for a broad spectrum of the general public to experience the performance and attractiveness of FCVs.
A brief explanation is given here of the 2005 X-TRAIL FCV model. Among other new technologies, this FCV is fitted with a new fuel cell stack that was developed in-house at Nissan. Compared with the previous 2003 FCV model, the vehicle performance of the 2005 X-TRAIL FCV has been markedly improved, particularly its driving range. In terms of practicality under ordinary operating environments, this FCV attains a level of performance comparable to that of conventional vehicles powered by an internal combustion engine (ICE).
Specifications of ,
Fig. 3. 2005 X-TRAIL FCV model [2]
2005 X-TRAIL FCV Specification
The major specifications of the 2005 X-TRAIL FCV models are given in Table 1. Previously, the following technologies were respectively accumulated at Nissan in the process of developing the Hypermini electric vehicle, Tino Hybrid vehicle and the AD Van CNGV that operates on compressed natural gas:
(1) a high-voltage traction motor and a lithiumion-battery system,
(2) an energy control system, and
(3) a high-pressure fuel storage system.
The 2003 and 2005 FCV models were developed on the basis of these core technologies. The 2005 model is virtually identical to the 2003 model in appearance, except for the body color, and in dimensions, except for the vehicle height. (The overall height of the 2005 FCV model is lower by 55 mm owing to the discontinuation of the rear spoiler.) However, the vehicle weight has been significantly reduced by 170 kg from 1,960 kg to 1,790 kg (1,860 kg for the vehicle with the 70 MPa hydrogen cylinder).
As will be explained later, the weight reduction was achieved by reducing the weight of the newly developed fuel cell stack and the weight of the electric system.
Table 1
X-TRAIL FCV [1]
Vehicle Length/Width/Height (mm) 4485/1770/1745
Weight (kg) 1790(1860)
Seating capacity 5
Top speed (km/h) 150
Crusing range (km) Over 370 (over 500)
Motor Type Coaxial motor integrated with reduction gear
Max. power (kW) 90
Max. torque (Nm) 280
Fuel cell stack Fuel cell Polymer electrolyte type
Max. power (kW) 90
Supplier Developed by Nissan
Battery Type Compact Lithium-ion Battery (Laminated type)
Fueling system Fuel type Compressed hydrogen gas
Max. pressure (MPa) 35 (70)
Major Technologies Vehicle Packaging The in-vehicle layout of the principal components is shown in Fig. 4. The fuel cell stack and high-pressure hydrogen storage cylinder are located under the floor, and the traction motor, inverter and other electrical system components are located in the engine compartment. Nissan's Compact Lithium-ion Battery was located behind the rear seatback on the 2003 FCV model, but it is positioned under the luggage area floor on the 2005 FCV model. As a result, the luggage area length of the 2005 FCV model has been increased by approximately 400 mm for improved practicality. In addition, the diameter of the 35 MPa high-pressure hydrogen cylinder has been reduced by approximately 50 mm compared with the cylinder of the 2003 FCV model. That smaller size allows a lower sitting height for the rear-seat passengers, which improves ease of ingress/egress and rear-seat habitability. The smaller diameter of the 35 MPa hydrogen cylinder reduces the onboard hydrogen storage capacity by 20% compared with the 2003 FCV model. However, the driving range of the 2005 FCV model fitted with the 35 MPa cylinder has been extended by 20 km over that of the 2003 FCV model as a result of improving the efficiency of the fuel cell stack and reducing the vehicle weight, among other improvements.
Fig. 5. In-house developed fuel cell stack [2]
70 MPa high-pressure hydrogen cylinder Fig. 6 is a photograph of the 70 MPa hydrogen cylinder that was newly developed for the 2005 FCV model in a joint project with Dyneteck Industries Ltd. of Canada. This cylinder raises the charging pressure twofold over the previous 35 MPa high-pressure cylinder. That increase boosts the hydrogen storage capacity by approximately 30% in the same outer volume compared with a pressure of 35 MPa. The new cylinder is made of an inner aluminum alloy liner and a carbon-fiber-reinforced outer shell. It has been certified by the High Pressure Gas Safety Institute of Japan (KHK) as a compressed hydrogen automotive fuel cylinder.
Inverter Compact Lithium-ion Battery
Fig. 4. Layout of major components [2]
Component Technologies Fuel cell stack Fig. 5 is a photograph of the new fuel cell stack developed in-house at Nissan. This newly developed fuel cell stack adopts a thin carbon separator that reduces the volume of the stack to approximately 60% of that of the fuel cell stack on the 2003 FCV model, yet maximum power output has been improved to 90 kW compared with 63 kW for the previous model. That improvement was achieved by improving the polymer electrolyte membrane and optimizing the operating procedure. However, durability and other attributes are still not sufficient compared with an IC engine and will require further improvement in the future.
Fig. 6. 70 MPa hydrogen cylinder [2]
In order to raise the charging pressure to 70 MPa, it was necessary to develop anew all the constituent parts of the fueling system, including the high-pressure hydrogen cylinder, hydrogen cutoff valve and pressure regulator valve, among others. That made it possible to increase the hydrogen storage capacity while ensuring reliability against hydrogen leakage. In addition, the in-vehicle layout of the system was also improved by positioning an in-tank valve at the mouth of the cylinder. This in-tank valve houses the cutoff valve, safety valve, hydrogen temperature sensor, reverse-flow check valve of the hydrogen charging system and the pressure-reducing valve of the hydrogen supply system.
A prototype FCV model fitted with the 70 MPa high-pressure hydrogen cylinder was tested on public roads in Canada in February 2006. Validation tests were
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
conducted concerning the effect of a temperature rise inside the cylinder during rapid charging of hydrogen and the effect of a temperature drop due to a rapid discharge of hydrogen under low ambient temperatures.
Electrical system Fig. 7 is a photograph of the traction motor fitted on the 2005 FCV model, and Fig. 8 is a photograph of Nissan's Compact Lithium-ion Battery.
Fig. 7. Traction motor [2]
2005 X-TRAIL FCV Vehicle Performance Driving range Fig. 9 shows the improvements achieved in the driving range of Nissan's FCVs in recent years. The driving range is one key issue for the practical use of FCVs. The 2005 FCV model was developed with the aim of achieving a driving range of at least 500 km under Japan's 10-15 emission test mode. Toward that end, the 70 MPa high-pressure hydrogen cylinder was newly developed to increase the hydrogen storage capacity by 12% compared with that for the 2003 FCV model (35 MPa). Additionally, the efficiencies of the fuel cell stack and electrical system were enhanced, the vehicle weight was substantially reduced, and the aerodynamic performance was also improved. As the combined result of these improvements, the driving range of the 2005 FCV model was extended to 500 km, compared with 350 km for the 2003 FCV model. However, even though the 2005 FCV model attains a driving range of 500 km, that is only the minimum level provided by conventional ICE vehicles. It will be necessary to improve the driving range further in future work. To accomplish that, it will be necessary to increase the quantity of hydrogen storable onboard the vehicle. With the current high-pressure hydrogen storage system, it is difficult to increase the storage capacity further because of vehicle layout limitations. Therefore, it will be necessary to develop a new system with higher hydrogen storage efficiency.
Fig. 8. Compact Li-ion Battery [2]
The weight of the traction motor was reduced by 5% from that of the motor used on the 2003 FCV model, while its maximum power output was increased by 5 kW. A coaxial reduction gear, like that used on the 2003 FCV model, was adopted to secure high space utilization efficiency. By using magnesium for the case, the weight was reduced by 5% compared with that of the 2003 FCV model. The Power Delivery Module (PDM) mounted in the engine compartment comprises an inverter and a DC/DC converter and functions to provide electric power from the fuel cell stack and Compact Lithium-ion Battery to the traction motor and other components. The PDM developed for the 2005 FCV model is 30% smaller than the unit used on the 2003 FCV model as a result of improving the layout of its internal parts and the cooling system.
Like the 2003 FCV model, the 2005 FCV model is fitted with Nissan's Compact Lithium-ion Battery consisting of laminated cells with high energy density. The controller that monitors the cell voltage was reduced in size and modularized, making it possible to mount the battery under the luggage area floor.
Fig. 9. FCV driving range improvement [2]
Power Performance Fig. 10 shows the improvements achieved in 0-100 km/h acceleration performance of Nissan's FCVs in recent years. The 2005 FCV model accelerates from 0 to 100 km/h in 14 sec., compared with 20 sec. for the 2003 FCV model, as a result of improving the power output of the fuel cell stack and traction motor and significantly reducing the vehicle weight. This improved acceleration performance compares favorably with that of conventional ICE vehicles. Additionally, the 2005 FCV model attains a top speed of 150 km/h, up from 145 km/h for the 2003 FCV model. These figures indicate that its power performance has reached a level that is sufficient for practical use.
Fig. 10. FCV acceleration improvement1
Crash performance The 2005 FCV model was developed to comply with the road vehicle safety standards that have been enforced on FCVs in Japan since March 31, 2005. Fig. 11 shows scenes of the frontal and rare-end collision tests that were conducted on prototype models in-house. The results of these tests confirmed that the 2005 FCV model meets not only the specified occupant injury criteria but also the road vehicle safety standards established for hydrogen safety and high-voltage safety.
Fig. 11. FCV crash performance [2]
Remaining Four Issues
There are four major issues that must still be addressed in order to put FCVs on the market: (1) to prevent performance decay, (2) to increase power density, (3) to improve subzero start-up capability and (4) to develop technologies for reducing costs. The present status of our efforts to resolve these issues is described below, based on the knowledge and insights gained through our R&D activities to date.
(1) Performance Decay
Method of etsimating FC stack performance degradation
This section outlines the method used to estimate the performance degradation of the FC stack on the X-TRAIL FCV. This method of estimating performance degradation was developed before an actual vehicle was built in order to make necessary estimates of stack durability and improve durability efficiently. Trying to investigate FC stack durability by building physical models and conducting durability tests and studies over a number of years would be time-consuming and costly. That approach was deemed to be an inefficient way of developing FC stack durability for FCVs.
Main cause of FC stack performace degradation The main causes of FC stack degradation in vehicle use are thought to be carbon corrosion at start-up, platinum (Pt) dissolution during load cycling and Pt dissolution and electrolyte membrane degradation during low current densities (high potential) such as when a vehicle is idling (Table 2) [3]. These causes have been reported in the literature and discussed in conference presentations [4, 5]. There are other possible causes of stack degradation besides those listed in the table, but the results of analyses done to date have not shown substantial degradation, so it is presumed that the influence of other factors is minimal.
Table 2
Fuel cell operating modes and major types of degradation [3]
Operating mode Degradation Main causes
Start-up Cathode catalyst surface area loss Cathodic reactant gas diffusion deterioration Cathode carbon support corrosion by high potential
Load cycling Cathode catalyst surface area loss Cathode catalyst dissolution by potential cycling
Idling (Iow current) Cathode catalyst surface area loss Membrane proton conductivity loss Cathode catalyst dissolution by high potential Chemical decomposition by peroxide (radical) attack Cathode catalyst poisoning by membrane fragments
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Carbon corrosion At start-up, both the anode gas channels and the cathode gas channels in the stack are filled with air (oxygen). When hydrogen is supplied to the anode in this condition, electromotive force occurs in the areas receiving hydrogen, but in the areas that do not receive hydrogen, the electromotive force produces a high electric potential on the cathode side. As shown in Fig. 12, this mechanism gives rise to degradation in the form of carbon corrosion [4-6]. Carbon corrosion causes the catalyst layer to collapse, and simultaneously the supported Pt particles aggregate, resulting in a decline in performance.
Fig. 12. Degradation mechanism at FC stack start-up [3]
Pt dissolution During load cycling and idling operation, changes in the electric potential and the holding of a high potential oxidize and reduce Pt. As a result, the surface of some Pt particles dissolves to produce an ion state (Fig. 13). The dissolution of Pt reduces the Pt surface area compared with the initial condition, which causes performance to decline [4, 7-14].
Electrolyte Membrane Degradation It has been estimated that electrolyte membrane degradation during low-current densities (high potential) causes performance to decline through the following mechanism [4, 15-17]. Hydrogen peroxides produced inside the electrolyte membrane decompose to form radicals that attack the membrane, thereby worsening the proton conductivity of the membrane, resulting in degradation of performance. However, it is known that serious degradation occurs under the harsh conditions of
high temperature and low humidity [15-22]. It is also known that FC stack performance declines due to catalyst poisoning by the decomposition products resulting from electrolyte membrane degradation [23]. Under ordinary vehicle control conditions, the influence on performance degradation is not so large, as will be explained later.
Separate Estimation of Degradation under Each Operating Mode Estimation of degradation for start/stop mode
Degradation that occurs during FC stack start-up and shut-down can be divided between these two processes. At start-up, the composition of the gases in the anode and cathode and the start-up control procedure cause the electric potential in the corroded part of the cathode to change, which varies the severity of the degradation that occurs. Assuming that the same start-up control procedure is applied every time, the oxygen concentration in the anode and cathode probably causes a pronounced difference in the degree of degradation. Even if all the oxygen in the cathode has been consumed at the time of shut-down, air permeates and diffuses from the gas seals of the FC stack and the piping at the inlet and outlet of the cathode. As a result, the inside of the cathode gradually changes to an air atmosphere. The inside of the anode also becomes an air atmosphere before long owing to the permeation of air (oxygen) from the cathode through the electrolyte membrane. In other words, degradation at start-up increases in proportion to the length of elapsed time since the last shut-down. For that reason, when estimating degradation at start-up in vehicle applications, it is necessary to ascertain the oxygen concentration in the FC stack at the time a vehicle is started.
The oxygen concentration in the anode and cathode was found in this study using the length of time between shut-down and start-up (referred to here as standing time) as the parameter. A test rig was used to conduct a start-up durability test under a condition with an air atmosphere in the anode and cathode. The amount of component degradation that occurred under various oxygen concentrations (0-21%) was also found in separate tests. The data thus obtained were used to estimate degradation sensitivity to the oxygen concentration. A frequency distribution of the standing time (from stop to start) for ordinary ICE vehicles was found and used in estimating start-up degradation.
The rate of degradation at shut-down was calculated in repeated start-stop cycles conducted with a test rig. It was assumed that loading cycles are applied to the FC stack when it is shut down under the same conditions every time.
In line with the preceding discussion, the following data are needed to estimate the degradation of FC stack performance at start-up and shut-down.
• Durability bench test data for repeated start/stop operation during a short period of time.
• Durability bench test data for repeated cycles in which the FC stack is completely changed to an air atmosphere in the interval from shut-down to start-up.
• The degree of start-up degradation relative to the oxygen concentration in the stack.
• The oxygen concentration in the FC stack relative to the standing time (from shut-down to start-up).
• A frequency distribution of the standing time of ICE vehicles.
Degradation estimation for load cycling mode
During FCV operation, the output of the FC stack changes every time the driver depresses or lets up on the accelerator pedal, causing the current and voltage of the stack to change. This produces potential cycling in the cathode catalyst layer that gives rise to degradation of performance. Three load patterns were defined for the FC stack-low load, medium load and high load cycles-as shown in Fig. 14, and durability bench tests were conducted under each load pattern. Degradation that occurs under load cycling was then estimated taking into account the number of times and the duration frequency of the three load patterns, based on the output frequency of ICE vehicles.
Pattern 3
Pattern 2
O
Pattern 1
Time
Fig. 14. Durability patterns defined for load cycling [3]
To recapitulate the preceding analysis, the following data are needed to estimate the degradation of FC stack performance during load cycling.
• Durability bench test data for low, medium and high load cycles.
• The distribution of the output frequency of ICE vehicles.
Degradation estimation for idling (low current) mode
As noted earlier, electrolyte membrane degradation and dissolution of the Pt catalyst layer are assumed to be the two principal causes of FC stack performance degradation during idling operation. It is thought that Pt dissolution causes a larger decline in performance than electrolyte membrane degradation. This is attributed to the fact that the FC stack on FCVs does not constantly operate at high temperature nor does it constantly operate under a low current condition. Accordingly, in estimating degradation of performance for idling
operation, bench tests were conducted under a condition where a low current density was continued for a certain length of time, followed by a larger current so as to change the electric potential. Degradation during idling was then estimated by taking into account the idling time of ordinary ICE vehicles based on their output frequency.
To recapitulate the preceding discussion, the following data are needed to estimate the degradation of FC stack performance under idling operation.
• Durability bench test data for load cycling that includes low current densities.
• The distribution of the output frequency of ICE vehicles.
Durability of Onboard FC Stack Estimated from Durability Bench Test Data
As explained in the preceding section, the loads imposed on an FCV during real-world driving were estimated from the manner in which ordinary ICE vehicles are operated. Degradation of performance under each operating mode was then simply summed to estimate FC stack degradation. Fig. 15 shows the ratio of the estimated degradation among the three operating modes.
Load cycling
Idling
□ Start/stop cycles
□ Idling
□ Load cycling
Start/stop cycles
Fig. 15. Ratio of estimated FC stack degradation by operating mode [3]
The results in the graph show that the start-stop mode accounted for 44% of the total degradation, representing an exceptionally large proportion. Both the load cycling mode and the idling mode showed approximately the same degree of degradation at around 28%. While degradation in the load cycling mode is large per unit of time, cumulative idling time is longer, which is probably why the two modes showed approximately the same values. Degradation can also be divided between that which occurs during start-up/shut-down and that which occurs during power generation (load cycling + idling). Degradation during start-up/shut-down accounted for 44% of the total and degradation during power generation accounted for 56%, indicating approximately the same relative degree of degradation.
While these results were estimated for the FC stack used on the X-TRAIL FCV, this method of estimating
performance degradation can also be applied to other FC stacks by taking into account the control procedure at start-up/shut-down, the control procedure during power generation and other relevant conditions.
Stack Performance Degradation Calculated from Vehicle Data Trend for decline in vehicle performance To date, the latest model of the X-TRAIL FCV has been driven on public roads for over two years, or in some cases, for three years in Japan and overseas combined, including use at numerous events in North America and Europe. The vehicle system and the FC stack both continue to display high levels of reliability.
Fig. 16 shows the change in FC stack performance as a function of the stack operating time for one vehicle that has a long cumulative operating time in Japan. FC stack voltage is shown along the vertical axis in relation to the FC stack operating time along the horizontal axis. These FC stack voltage data were extracted only for high load conditions above a certain specified current value. It is seen that FC stack voltage, i.e., performance, has gradually decreased as the cumulative operating time has increased. In order to quantify the degradation of each FC stack performance parameter, the FC stack voltage was averaged for every 100 hours of operating time. The values were then plotted to eliminate the influences of different operating environments, and linear approximations were made to calculate the performance degradation. In making the linear approximations, data that singularly deviated from a straight line were omitted. This method was used to express the levels of degradation for various FC stacks.
500 1000 1500
FC stack operating time [hours]
2000
Fig. 16. Stack voltage as a function of FC stack operating time during FCV driving in Japan [3]
Change in FC stack performance on experimental FCVs Fig. 17 shows the degradation of FC stack performance in relation to the FCV operation period. It is seen that performance degradation tends to increase with a longer operation period, but there are also FC stacks that show a slight decline in performance even though they have been in operation for more than a year. These data refer presumably to vehicles that have been driven rather infrequently since put in operation.
Degradation of FC stack performance as a function of the stack operating time is shown in Fig. 18. The data indicate a large decline in performance with a longer operating time. Excluding some of the data plots, the data for many of the vehicles indicate that performance tends to decline in proportion to the FC stack operating time.
Fig. 19 shows the degradation of FC stack performance as a function of FCV mileage. While degradation of performance tends to increase with longer mileage, there are also many stacks that show degraded performance even though the mileage of the FCVs is relatively short. This suggests that the degradation of FC stack performance does not correlate strongly with vehicle mileage.
t <u CL
0.5 1 1.5 2 FCV operation period [years]
2.5
Fig. 17. Performance degradation as a function of FCV operation period [3]
t <u CL
500 1000 1500
FC stack operating time [hours]
2000
Fig. 18. Performance degradation as a function of FC stack operating time [3]
<u CL
0 20000 40000 60000 80000
FCV mileage [km]
Fig. 19. Performance degradation as a function of FCV mileage [3]
3
0
0
Degradation of FC stack performance as a function of the number of start/stop cycles is shown in Fig. 20. Larger performance degradation occurs as the number of start/stop cycles increases.
ro E o t <D CL
□ Load cycling (Estimation)
i-L □ Idling (Estimation)
□ Start/stop cycles (Estimation)
-rv O Vehicle data
o o
m
Japan #1 Japan #2 US #1
US #2
Fig. 21. Comparison of estimated and actual performance degradation for different operating modes [3]
0 1000 2000 3000 4000 5000 6000 Number of start/stop cycles
Fig. 20. Performance degradation as a function of number of start/stop cycles [3]
The causes of FC stack performance degradation were investigated on the basis of data collected in actual driving. However, as the preceding results indicate, there was no one single parameter that showed an exceptionally strong correlation with performance degradation. This implies that the degradation of FC stack performance is caused by multiple factors.
Comparison of Vehicle Data and Performance Estimations
The preceding section described the estimations of FC stack performance degradation calculated with the proposed method and the levels of performance degradation found for FC stacks used on experimental FCVs. This section compares the estimations of FC stack performance degradation with the actual performance degradation results found for experimental vehicles in order to validate the accuracy of the proposed method for estimating performance degradation.
Comparison of FC stack degradation estimations and actual vehicle data
The rate of FC stack performance degradation was calculated from the driving data recorded for the stacks used on experimental vehicles. The loads applied to those FC stacks were calculated from the number of start/stop cycles and the power generation time. Based on the load data, the proposed method for estimating FC stack performance degradation was used to calculate the rate of performance degradation. A comparison was then made between the two sets of performance degradation rates.
Fig. 21 compares the performance degradation rates based on actual vehicle data and the estimated rates for the three operating modes. These data refer to four FCVs that have accumulated relatively long operating hours either in Japan or in the U.S.
The performance degradation rates calculated from the vehicle data and the estimated performance degradation rates show good agreement. The total number of start/stop cycles did not differ substantially from the initial prediction, but the frequency of standing time was different from the assumed value and there were fewer starts following a long standing time. Consequently, the proportion of degradation attributed to start/stop cycles was lower for each vehicle than the estimated values. The total FC stack operating time did not differ appreciably from the assumed value, but the assumed frequency of the three load levels differed because each vehicle was operated in a different driving environment. As a result, some vehicles showed larger degradation during idling while others showed larger degradation during load cycling, so the proportion of degradation ascribable to each operating mode differed considerably among the vehicles.
iT ja'
□ -'B
□ X
D
□ X
s' 1% 1-1
Degradation (Estimation)
Fig. 22. Comparison of estimated and actual performance degradation for vehicles driven more than 18 months [3]
Fig. 22 compares the performance degradation rates based on vehicle data and the estimated values for vehicles that have been driven for more than 18 months and for which their detailed driving logs are known. For nearly all of the vehicles, the estimated values agree well with the performance degradation rates based on the vehicle data.
The preceding results clearly show that the performance degradation levels estimated with the
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
proposed method coincided well with the levels of performance degradation found for FC stacks having a certain cumulative operating time. This verifies that the method proposed here for estimating FC stack performance degradation provides good accuracy.
Performance Evaluation and Disassembly Examination of Used Stack MEAs The membrane electrode assembly (MEA) was randomly selected from some used FC stacks for conducting a single-cell performance evaluation and disassembly examination. A check was made to confirm the occurrence of carbon corrosion and Pt dissolution, two previously predicted causes of degradation, and also the presence of electrolyte membrane degradation.
Evaluation of I-V Characteristics Power generation tests were conducted on a single cell of used and new MEAs under conditions simulating those of actual FCV operation. The results are shown in Fig. 23 in terms of the I-V characteristics.
No difference was observed between the new and used MEAs under a no-load condition, but a difference in voltage occurred during operation. The voltage difference was not proportional to the current value at each level of current density, which implies that there was virtually no influence from increased resistance of the electrolyte membrane. The decline in voltage on the low current region is a tendency characteristic of a rise in the activation overpotential of the cathode, which presumably indicates a reduction of the Pt surface area of the cathode catalyst layer.
Current density
Fig. 23. Results of power generation performance tests (I-V characteristics) [3]
Electrochemical Diagnostic Tests Electrochemical diagnostic tests were conducted to investigate the decline in the I-V characteristics in more detail. The electrochemical surface area (ECA) of the anode and cathode was calculated by cyclic voltammetry (CV). The results are shown in Fig. 24 in relation to the initial values of a new MEA.
Compared with the initial values, the ECA of the anode was reduced by approximately one-half and that of the cathode to about one-third. The activity of the hydrogen oxidation reaction (HOR) that occurs at the
anode is much higher than that of the oxygen reduction reaction (ORR) that occurs at the cathode. Accordingly, the ECA of the anode does not influence performance greatly, but a loss of the cathode ECA causes degradation of performance. The theoretical voltage drop calculated from the loss of the cathode ECA coincided with the decline in performance, indicating that the reduced ECA of the cathode was the principal reason for the voltage drop.
1GG
о4 9G
e SG
га > 7G
6G
'с 5G
to
e 4G
> 3G
га
2G
AC 1G
E G
и
□ new MEA
□ used MEA
Anode
Cathode
Fig. 24. Results of electrochemical diagnostic tests [3]
Post-test analysis of MEAs A post-test analysis was made of new and used MEAs to check for the presence of carbon corrosion, Pt dissolution/aggregation and electrolyte membrane degradation. A scanning electron microscope (SEM) was used to observe the cathode catalyst layer. Typical SEM images obtained are shown in Fig. 25.
Fig. 25. SEM images of cathode catalyst layer [З]
In the SEM image of the new MEA, the fine structure of the carbon particles of the catalyst layer can be observed along with the presence of many pores. In contrast, the images of the MEA from a used FC stack show that the carbon particle structure has been destroyed and the number of pores reduced. It is seen that destruction of the carbon particle structure is more severe at the anode outlet than at the anode inlet. These results are consistent with the findings reported for an analysis of MEAs subjected to repeated start-stop tests [2]. They clearly show that corrosion of the carbon support of the cathode catalyst layer occurs at start-up in FC stacks used on vehicles.
New and used MEAs observed in their entirety by electron probe microanalysis (EPMA) are shown in Fig. 26. No Pt was observed inside the electrolyte membrane of the new MEA, but Pt was observed inside the electrolyte membrane of the used MEA at a location near the cathode catalyst layer. Presumably, Pt of the cathode catalyst layer dissolved during potential cycling and idling operation to form an ion state and was reduced and deposited inside the electrolyte membrane [2]. This indicates a reduction in the amount of Pt on the cathode catalyst layer. However, the amount of Pt inside the electrolyte membrane was markedly smaller than the Pt loading of the cathode catalyst layer. Accordingly, the amount of Pt inside the electrolyte membrane alone would not explain the reduction of the cathode ECA to one-third of its initial value. As another reason for loss of the ECA, it is known that dissolved Pt is re-deposited over the Pt of the cathode catalyst layer, causing the Pt particle size to increase and reducing the Pt surface area [14]. The EPMA results suggest that Pt dissolution and aggregation occur during load cycling and idling operation in the FC stack on FCVs.
New MEA
Cathode catalyst layer
It is also seen in Fig. 26 that the thickness of the electrolyte membrane was the same as the initial value and that there were no cracks in the membrane. It is inferred from these observations that serious degradation of the electrolyte membrane did not occur.
In the process of developing the proposed method for estimating the degradation of FC stack performance, it was assumed that carbon corrosion and Pt dissolution were among the principal causes of performance degradation of the FC stack for vehicle use. The preceding investigation of MEAs actually used on experimental FCVs verified those assumptions. The results also showed that the influence of electrolyte membrane degradation was minimal. Since the estimations of FC stack performance degradation agreed well with the actual levels of degradation found for used FC stacks, it is inferred that carbon corrosion and Pt dissolution are the principal causes of the degradation phenomena that occur in the FC stack.
Durability Improvement
The durability estimations made for the FC stack used on the 2005 X-TRAIL FCV were shown to be sufficiently accurate. That information was then put to use in developing the durability of the new FC stack which will be explained later. As the results described above have indicated, there is no single factor governing degradation, but the evaluations made it clear that it was necessary to reduce carbon corrosion and Pt dissolution. In addition, from the standpoint of cost, it was necessary to reduce the loadings of Pt and other precious metals. That presented an exceptionally tough issue with respect achieving an acceptable balance with other performance parameters.
In developing the new FC stack, degradation due to carbon corrosion at start-up/shut-down was addressed through the system design. Material improvements were made to deal with Pt dissolution. With regard to the electrolyte membrane, it was confirmed that there was no increase in the amount of the hydrogen crossover through the membrane. As a result of these improvements, performance degradation of the new FC stack has been successfully reduced compared with the 2005 model (Fig. 27).
Anode catalyst layer
Used MEA
Cathode catalyst layer
Anode catalyst layer
Fig. 26. EPMA images of Pt catalyst layer of new and used MEAs [3]
in LU
<D O
□ Others*
□ Load cycling
□ Idling
□ Start/stop cycles * Others:
Sub-zero startup, etc.
2005 Model New stack
Fig. 27. Comparison of FC stack durability [3]
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
The main measures taken to prevent the degradation of stack performance are outlined below.
• The system has been designed to maintain a low oxygen concentration in the FC stack, which is a principal cause of degradation due to carbon corrosion at start-up, thereby improving stack durability.
• The structure of the cathode catalyst layer was improved to address degradation due to Pt dissolution during load cycling. This improvement enhances durability while maintaining high performance.
• The degradation estimations made in this study showed that virtually no electrolyte membrane degradation occurred due to idling operation (low current density). With an eye toward further system improvements in the future, the material of the electrolyte membrane was improved to increase the durability of the FC stack.
Future Work
It is planned to confirm and further improve the accuracy of the degradation estimation method by accumulating more real-world data though the continuation of FCV driving tests. Investigations will be undertaken to check for new forms of unexpected degradation that might develop over longer periods of FC stack use on vehicles. It is also planned to investigate forms of degradation that can occur suddenly, rather than causing a decline in performance proportional to the operating time. This will include examining increases in hydrogen crossover through the electrolyte membrane, degradation due to fuel starvation at the anode and degradation originating in the stack structure.
It should be noted that carbon corrosion and Pt dissolution were revealed to be the major causes of FC stack performance decay because the results indicated that the electrolyte membrane used in the test FCVs suffered little degradation. Based on these results, we are proceeding with development work aimed at further suppressing the causes of stack performance degradation. In addition to developing more durable catalysts and other materials, this work also includes efforts to elucidate the operating conditions required by FC systems in order to inhibit stack performance decay.
(2) Power Density
An important condition for implementing an FC system in a vehicle is to increase the power density of the system by reducing its size and increasing its power output. This section describes measures for increasing power density, focusing on the FC stack.
Two basic approaches can be cited for achieving higher power density. One is to downsize the components of the FC stack, and the other is to improve power output performance. At Nissan, we improved the first-generation FC stack used on the
2005 model X-TRAIL and announced an improved second-generation stack in August 2008 [3]. Shown in Fig. 28, this new FC stack achieves approximately a twofold improvement in power density as a result of incorporating the two approaches mentioned here. Development work is now under way to build a next-generation stack with the aim of achieving outstanding power density of 2.5 kW/L, enabling the unit to be adopted in parallel across various vehicle model lines (Fig. 29).
Specifications
Max. Power 130 kW
Volume 68 L
Weight 86 kg
Fig. 28. 2GGS Model FC Stack [З]
_i 2.5
I
I 2
«
n e
a 1.5
0.5
Next (Generation
2008 Moeel
- Л
/
--
Target
2005 Model /_
2000 2002 2004 2006 2008 2010 2012 Fig. 29. Roadmap of FC Stack Power Density [1]
Metal Separators One possible way of downsizing the FC stack is to reduce the size of the layered stack structure and other structural components. For the 2008 model FC stack, metal separators were adopted for the specific purpose of downsizing the layered structure (Fig. 30). The metal separators are fabricated of thin metal plates that are shaped by press forming to provide the gas and coolant channels. Compared with the previous carbon-plastic separators (Fig. 31), the thickness of the separator plates can be reduced by more than one-half while providing high mechanical strength. As a result, the cell pitch was reduced to two-thirds of the previous pitch, which, together with the downsizing of non-layered parts, made it possible to reduce the volume of the 2008 model FC stack by approximately 25% compared with the previous stack.
H2-Coolant-Air
Cell pitch
L
MEA
Carbon separator Fig. 30. Configuration of 2005 model [24]
H2
Coolant Air
MEA
3 -
m S
Cell pitch
2/3L
Metal separator Fig. 31. Configuration of 2008 model [24]
When metal separators are used, it is necessary to ensure both high corrosion resistance and high conductivity. The metal separators adopted for the 2008 model stack are given a surface treatment that satisfies this requirement. The surface treatment material was selected from among several candidate materials on the basis of a conductivity test and a corrosion evaluation test conducted on test pieces under conditions simulating the corrosive environment inside the cells [25, 26]. The results of conductivity tests conducted on the selected surface treatment material are shown in Fig. 32 [27]. The adoption of this surface treatment has substantially reduced contact resistance between the separators and the gas diffusion layer (GDL) and also between the separators.
Carbon
Metal
Carbon
Metal
Separator/ GDL
Separator/ Separator
Fig. 32. Comparison of contact resistance between carbon and metal separators [27]
A cell with metal separators and one with the previous carbon-plastic separators, for which there is no corrosion problem, were subjected to a durability test using a small single cell. The tests were conducted under accelerated conditions of high potential and low humidity, which were designed to be severe for corrosion of the metal separators [28]. The results are plotted in Fig. 33. The effects of corrosion include electrolyte membrane degradation caused by metal ion precipitation and a decline in FC stack performance due to increased conduction resistance resulting from surface oxidation. The results show that the metal separator cell tended to display the same level of voltage drop as the cell with carbon-plastic separators. This indicates that there was no increase in conduction resistance due to corrosion of the metal separators. In addition, the amount of metal ions contained in liquid droplets collected at the gas outlet during the test was less than 6 ppb at both electrodes for the metal separator cell. It was comfirmed that the amount was not significant for accelerating degradation of the electrolyte membrane [29]. These results imply that metal precipitation did not cause any significant performance degradation.
Fig. 33. Comparison of contact resistance between carbon and metal separators [27]
Enhancement of Power Generation Performance through MEA Improvements
As mentioned earlier, besides downsizing, another approach to increasing the power density of the FC stack is to improve power output performance. That can be effectively accomplished by reducing polarization during power generation. In order to reduce polarization, it is important to design the stack on the basis of a good understanding of the phenomena involved in the transport of protons, water, heat, electrons, air, hydrogen and other elements inside the MEA. For the MEA of the 2008 model stack, the molecular structure of the polymer electrolyte membrane was modified so as to increase the water content of the membrane and thereby enhance proton conductivity, as shown in Fig. 34 [24]. That had the effect of reducing the resistance loss occurring in the electrolyte membrane. In addition to that improvement,
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
the electrolyte membrane was also improved so as to increase its capability for back diffusion of water through the membrane. This refers to the membrane's ability to transfer water produced at the cathode electrode to the anode electrode. The capability for back diffusion of water is a key characteristic for preventing dehydration of the electrolyte membrane on the anode side. This is especially critical in the operating condition of high current density where the amount of water molecules dragged by electro-osmosis from the anode side to the cathode side increases.
■ 2005 Model FC Stack MEA -2008 Model FC Stack MEA
о -d
n o C n o
Рч
0 2 4 6 8 10 12 Water Content (k, [H2O]/[SO3-])
Fig. 34. Relationship between water content and proton conductivity in membrane [24]
o
>
C
У
У
20 mV
Fail
■ 2005 Model FC Stack MEA 2008 Model FC Stack MEA
20 40 60
Cathode Inlet RH (%)
100
Fig. 35. Cell voltage sensitivity to cathode humidification [24]
Fig. 35 shows the cell voltage as a function of the relative humidity at the cathode inlet of the 2008 model MEA that incorporates the above-mentioned improvements [24]. Compared with the results for the previous MEA, the 2008 model MEA shows high cell voltage at all levels of cathode humidification, with an especially large improvement seen for a low level of relative humidity at the cathode inlet. This indicates that the improved MEA can ensure proton conductivity even if the amount of water supplied by external
humidification is reduced. It should be noted that this improvement in cell voltage under low humidification conditions is an especially important performance attribute because it allows the external humidifier on the cathode side to be downsized, along with facilitating other improvements. As a result, it helps to reduce the size and cost of the overall FC system significantly.
(3) Subzero Start-up Capability
The subzero start-up capability of a fuel cell is an essential functional requirement in order to promote the use of FCVs over wide geographical areas. Vehicle manufacturers have been pushing ahead with R&D efforts to address this crucial issue. The subzero startup-capability of the newly developed 2008 model FC stack has been markedly improved. This section outlines the subzero start-up capability improvement.
Fuel Cell Cold Start Concept
The experiments conducted by Kagami et al. using a laboratory-scale single cell show that water produced at the cathode catalyst layer (CL) during subzero FC operation causes the ice fraction to grow in the CL [30]. This ice formation plugs CL pores, starving the reactant gases for the oxygen reduction reaction. Consequently, if the cell temperature at that time does not rise above 0 deg.C, the cell voltage drops drastically and the cell can no longer generate electricity. Kagami and his co-workers also discovered that a cell can continue to generate electricity if its own heat resulting from power generation is used to raise the temperature and transform the water produced to vapor which the reactant gas can remove. Oszcipok et al. also found that, in addition to the gas flow rate, the initial membrane water content prior to cell freezing is an important factor in FC cold start capability [31].
For the purpose of avoiding fuel cell shutdown due to ice formation, Mao et al. considered the water balance during subzero cell startup and reasoned that product water must either be diffused into the membrane or transported out into the gas diffusion layer (GDL) [32]. They developed an analytical model that took into account both water removal mechanisms and performed numerical calculations. Tajiri et al. used a laboratory-scale single cell to conduct subzero power generation experiments [33]. Their results show that cell resistance decreases accompanying power generation, indicating that the membrane actually absorbs product water generated at the cathode CL during power generation at subzero temperatures.
Accordingly, when a fuel cell is operated at subzero temperatures, power generation can be continued during the interval in which ice formation in CL pores is retarded through absorption of product water by the membrane or its transport by vapor to the GDL.
Cold startup of a fuel cell can be achieved during the period when power generation is continued without CL
0
pores becoming plugged with solid ice. That is accomplished through self-heating associated with power generation, which raises the cell temperature beyond the melting point of ice to ensure continued oxygen transport. As a result of recent technological advances, R&D work has now entered the phase of improving the sub-zero cold start capability of FCVs rather than simply ensuring subzero startup. Accordingly, it is necessary to discuss the cold start technologies related to cell performance during and after subzero startup. In other words, when considering the cold start performance of an FCV, what is important is how quickly high power output can be obtained when starting the vehicle from subfreezing temperatures.
An approach is explained here for determining the relative level of cold start capability. As mentioned above, during cold startup of a fuel cell, product water freezes and plugs CL pores, causing the cell voltage to drop and cell performance consequently declines. In other words, in order to improve the cold start capability of an FC stack, it is necessary to minimize the growth of the ice fraction in the CL -as much as possible- during cold startup.
In power generation tests of a fuel cell under isothermal conditions at subfreezing temperatures, Tajiri et al. observed that power generation was sustained to some extent even at subzero temperatures until the cell voltage declined sharply after a certain period of time had elapsed [33]. Mao et al. performed calculations with a multiphase model in an effort to explain that phenomenon [34]. Their calculations show that during cold startup under certain given conditions product water formed at the cathode CL is transported into the membrane and simultaneously ice forms in the cathode CL pores at a certain rate. This is determined by the balance between the water flux to the membrane and the rate of water production by power generation. For example, under a low load condition where the water production rate is slow, there is sufficient water flux to the membrane compared with the rate of water production. In this case, the ice fraction in the CL grows more slowly. After the membrane has absorbed product water to a certain extent, its water content increases, thus reducing the rate of water transport into the membrane, which is driven by the k gradient, and therefore water flux to the membrane declines. As a result, the ice formation rate in the CL increases to a corresponding extent. In the end, CL pores become plugged with ice, preventing the arrival of oxygen and the cell voltage consequently drops.
As explained by this model, even though the membrane has water uptake potential in the initial stage of cold startup, ice steadily forms and expands in the cathode CL owing to the water balance. Therefore, the more the load is increased in an effort to obtain higher power output, the steeper the cell voltage drop becomes and eventually the cell is unable to generate any electricity. This phenomenon characterizes the cold startup behavior of a fuel cell. One governing factor of
the relative cold start performance of the MEA is the rate of ice formation in the CL. It is inferred that cold start capability improves to the extent that ice accumulation in the CL is retarded.
It is virtually impossible to directly measure either quantitatively or qualitatively the amount of ice that forms in the CL during cold startup. By comparing the cumulative product water reported by Tajiri et al. as the output of the single cell used in their isothermal cold start tests, it is possible to make a qualitative comparison of the ice formation rate in the CL for different power generation conditions and MEA specifications [33]. For example, suppose a comparison is made of two MEAs under the same operating conditions. The one showing a longer continuous operational time and a larger amount of cumulative product water would signify a slower ice formation rate in the CL. As a result, it could be inferred that the cold start performance of this MEA is better.
As will be explained later, the MEA developed for the 2008 model FC stack was improved so that it produces more cumulative product water under the same operating conditions than the MEA of the 2005 model FC stack. This improved MEA contributed to the enhanced cold start capability achieved for the 2008 model FC stack.
Tajiri et al. reported that cumulative product water varied with temperature [35]. As shown in Fig. 36, they found that cumulative product water increased exponentially as the startup temperature approached 0 deg. C. One reason for this is that water diffusivity in the membrane increases with rising temperature, thus increasing water transport to the membrane. A further reason is that the amount of product water carried by the gas increases.
-30 -20 -)0 0
Start Up Temperature f C)
Fig. 36. Relation between product water and startup temperature [33]
In line with these characteristics, the ice formation rate in the CL can be suppressed by lowering the fuel cell thermal mass in order to quicken the temperature rise of the MEA and the cell. In other words, lowering the cell thermal mass is one effective way of improving cold start capability.
The concept formulated in this work for improving cold start capability is outlined below.
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
• Improvement of cold start capability (attainment of power output as quickly as possible following subzero startup) equals suppression of the ice formation rate in CL pores during cold startup/
The following two methods can be applied for suppressing the ice formation rate in CL pores.
• Increase the product water uptake potential and uptake rate into the membrane/
• Reduce the cell thermal mass/
The following sections describe the improvements made to the MEA developed for the 2008 model FC stack and the reduction of the cell thermal mass. The results of performance evaluations of a 16-cell short stack are also presented to illustrate the respective effects of these measures on cold start performance.
Cold Start Technoligies MEA Improvements Mao et al. used a multiphase, transient model to calculate ice formation in the CL of a polymer electrolyte fuel cell (PEFC) during cold startup [34]. In Fig. 37 the cumulative product water is plotted along the vertical axis in relation to the membrane water uptake potential (obtained by subtracting the initial membrane water content (X) at cold startup from X 14) along the horizontal axis. The pronounced sensitivity of the cumulative product water to current density ins noteworthy.
t—'-1—1—г
Start-up Temperature: -20 С Membrane Thickness: 30 jjm Purge Condition: Dry purge
^-sat'^O
Fig. 37. Effect of startup current density on cold-start performance [34]
In Fig. 38, the water content distribution following the cell voltage drop during isothermal cold operation is plotted along the vertical axis in relation to the position of the MEA cross section along the horizontal axis. These results indicate that the distribution and quantity of product water absorbed by the membrane differ depending on the current density during cold operation. Product water is concentrated more toward the anode side of the membrane at the lower current density.
Position (um)
Fig. 38. Water content profiles in the MEA at the end of isothermal cold start under different current densities [34]
The data indicate that cumulative product water during cold startup is closely related to water transport in the membrane. The results show that this is determined by the water balance between electro-osmotic drag (EOD), based on the amount of power generated, and the water generation rate per unit time, in relation to water diffusivity in the membrane and back diffusion related to the water content gradient (AX). This understanding provided a significant insight for elucidating the mechanisms involved in FC cold startup.
Similarly, Mao et al. examined whether cumulative product water differed for membrane samples of varying thickness [34]. They found virtually no difference in cumulative product water between a 30-^m membrane and a 45-^m membrane, as seen in the results in Fig. 39. The reason for that can be understood from the data in Fig. 40. The region of water uptake during cold startup is seen to be a thickness of around 20 ^m from the cathode CL for both the 30-^m and 45-^m membranes. The remaining portion of the membrane thickness does not provide a water storage function.
Fig. З9. Effect of membrane thickness on cold-start performance [34]
14
12
| 10
o
o
I—
0
1 6
-' 1 1 1 |
I 1 Anode 1 CL l L, Membrane 30 „m membrane / /' ' /' / 1 /1 / 1 1 j j I Cathodal / i / I CL \j / I /11
1 / 45 ur / 1 I / 1 I / 11 / 11 lj/rn^mbrane 1 1 -
1 " II
i 1 , , 1 , , , , 1 , , , , 1 , I 1 , 1 i . i i 1...... , r
10 20 30 40 Position (jjm)
50
60
Fig. 40. Water content profiles in the MEA at the end of isothermal cold start with varying membrane thickness [34]
These studies results suggest that the water uptake potential of the membrane, which substantially affects cold start performance, is more influenced by the water absorption rate than by the water storage capacity of the membrane.
Based on these numerical calculation results, we concluded that increasing water diffusivity in the membrane would be an effective way to improve the MEA for the purpose of enhancing cold start capability. Increasing water diffusivity in the membrane would also have the effect of improving performance at the normal operation temperature, especially under a condition of low humidification. Therefore, this approach would also be a convenient way of improving cell performance at normal operation temperature.
For the purpose of increasing water diffusivity in the membrane, the density of the sulfonic acid group in the polymer electrolyte membrane was increased in the MEA adopted for the 2008 model FC stack. While it is well known that this approach is effective in improving water diffusivity in the membrane, increasing the density of the sulfonic acid group also tends to reduce membrane durability. However, it was confirmed that improving water diffusivity in this way for the newly adopted MEA did not have any adverse effect on durability [27].
Various methods can be used to evaluate water diffusivity in an electrolyte membrane. One method is to measure the self-diffusion coefficient of the electrolyte membrane using nuclear magnetic resonance (NMR) [36]. A second method is to inject water vapor with different levels of water activity into both electrodes of the membrane (or to inject liquid water into one side) and then measure the quantity of diffused water [37]. A third method is to evaluate water diffusivity relatively by making a comparison of the limiting current density at which a higher load current can no longer be obtained, after inducing a sharp drop in the cell voltage by raising the load current under a condition of a sufficient supply
of reactant gases [38]. With this third method in particular, there is a marked decline in the proton conductivity (water content) of the electrolyte membrane on the anode side in the vicinity of the limiting current density. Measuring the level of the limiting current density can therefore be used as an index of the membrane's ability to return to the anode side the water transported with protons to the cathode side by electro-osmosis.
The limiting current density of the 2005 model FC stack MEA and that of the 2008 model FC stack MEA were evaluated in relation to cathode relative humidity and the results are shown in Fig. 41. It is seen that the limiting current density of the MEA adopted for the 2008 model FC stack was improved over the entire relative humidity range from low to high. These results experimentally confirm that water diffusivity from the cathode to the anode was improved in the 2008 model FC stack MEA over that of the 2005 model FC stack MEA by reducing the equivalent weight (EW) of the membrane.
Ö <u Q 'S
<u
O
J
20
40 60 Cathode RH (%)
80
100
Fig. 41. Comparison of limiting current density between 2005 model FC stack MEA and 2008 model FC stack MEA. As reactant gases, hydrogen was supplied at the anode and nitrogen at the cathode and the humidity was 20% at the anode [24]
However, care must be taken regarding data obtained by measuring the limiting current density. For one reason, this method evaluates the net water diffusivity between back diffusion and EOD, where the driving force is the water content difference in the cross-sectional direction of the membrane. Another reason is that the data are greatly influenced by heat, as was pointed out by Tabuchi et al. [38]. For example, the experimental results in Fig. 41 were measured using the same experimental environment, including the flow channels and jigs. However, since the lower-EW membrane of the 2008 model FC stack has higher performance and smaller overvoltage, it generates relatively less heat, so its temperature is that slightly lower. That could result in measurement of a higher limiting current density.
0
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
As explained here, making a detailed comparison of the water diffusivity and associated mechanisms of two MEAs requires advanced measurement techniques, including accurate measurement of the permeability, diffusivity and EOD of the membrane and compensation for the temperature due to differences in overvoltage. Accordingly, it is hoped that measurement methods and calculation models will be further improved in future studies.
The MEA adopted for the 2008 model FC stack also improves proton conductivity (Fig. 34) simultaneously with water diffusivity. This improvement is also beneficial for the cold start capability of FCVs in terms of how quickly high power output can be obtained.
The 2008 model FC stack MEA with a lower-EW membrane for improving water diffusivity between the cathode side and the anode side of the membrane was then subjected to an isothermal cold start test based on the experimental protocols reported by Tajiri et al., including equilibrium purge [33]. As the results shown in Figs. 42 and 43 indicate, the MEA of the 2008 model FC stack provides a longer operational time than that of the 2005 model FC stack.
1.2 1
0.8 0.6 0.4 0.2 0
>
й >
и
current density 40 mA/cm Jfc_SE_
-50
50
150 250 Time (s)
350
450
Fig. 42. Time to failure measured for various initial water contents under isothermal operation at -20 deg. C for 2005 Model MEA [24]
>
U
1.2 1
0.8 0.6 0.4 0.2 0
- - -*-X=9 -o-X=6 ^X=4.5 -
Х=9 :
1 Х=4.5 X=6¿
current density 40 mA/cm2 ¡i
-50
50
150 250 Time (S)
350
450
Fig. 43. Time to failure measured for various initial water contents under isothermal operation at -20 deg. C for 2008 model MEA [24]
Based on these results, the cumulative product water was then calculated from the current density during the interval of continuous power generation, and the results for the two MEAs are compared in Fig. 44. It is evident that the MEA of the 2008 model FC stack is capable of absorbing more product water. In other words, its water uptake rate into the membrane during cold startup is higher, which means that the ice formation rate in CL pores is retarded.
1.6 R 1.4
a
№ 1 2
£
S-h
3 1
0.8
0.6 0.4
A 2005 Model FC Stack MEA • 2008 Model FC Stack MEA
6
^sat -
10
12
Fig. 44. Comparison of product water during isothermal cold start operation on 40 mA/cm2 from -20 deg. C [24]
Various factors are intricately involved in the uptake of product water into the membrane during cold startup, including subzero EOD, water content and their temperature sensitivities, in addition to water diffusivity in the membrane. That complexity makes it exceptionally difficult to explain the experimental results quantitatively on the basis of a model. However, research efforts to achieve that explication have significant value in that it will then be possible to know which MEA materials affect cold start capability and to what extent. We are continuing such efforts as part of our activities to develop an MEA that further improves cold start capability.
The properties of the new MEA were expected to improve the cold start performance of the 2008 model FC stack, but the improvement achieved did not depend solely on the MEA properties. Rather, the following properties also contributed to improved cold start capability.
As pointed out by Oszcipok et al. and is also shown in Fig. 44, the initial membrane water content prior to cold startup affects the amount of water uptake into the MEA during startup from subzero temperatures [31]. It is clear that this greatly influences the ice formation rate in CL pores during cold startup. Accordingly, reducing the membrane water content as much as possible before cold startup while the FC system is still not operating can lead to improved cold start performance.
The performance of the MEA adopted for the 2008 model FC stack has been markedly improved in the normal operation temperature region under low cathode
relative humidity, as shown in Fig. 35 [27]. This improvement is attributed to the lower-EW membrane adopted for this MEA, as was explained earlier, and also to improvements made to the GDL and other components. As a result, it made it possible to reduce the amount of humidification provided by the water recovery device (WRD) in the power plant system of the FCV equipped with the 2008 model FC stack, allowing the WRD to be downsized by 40%.
In short, the membrane water content during ordinary operation can be reduced because the 2008 model FC stack can always operate under a condition of lower relative humidity at normal operation temperature than conventional stacks. As a result, the membrane water content can be reduced before cold start, thereby improving cold start capability.
Thermal Mass Reduction The configurations of the 2005 model FC stack and the 2008 model FC stack are shown in Figs. 30 and 31, respectively. The 2008 model FC stack was significantly downsized and its thermal mass was reduced by changing the separator material to metal from the previously used carbon.
The metal separator is fabricated from stamped thin metal plates that form the gas and cooling channels. Compared with the previous carbon separator, the thin metal plates provide ample mechanical strength and gas sealing performance, enabling the separator plate thickness to be reduced to less than one-half the previous dimension. That contributed substantially to the smaller stack size and reduced thermal mass.
The adoption of this separator fabricated from thin metal plates made it possible to reduce the thermal mass of the 2008 model FC stack by 32% compared with that of the 2005 model FC stack (Fig. 45).
- 32% reduction
^M 4,_
2005 Model 2008 Model Fig. 45. Thermal mass of 2005 & 2008 model FC stacks [24]
Experimental Results for FC Stack and Power Plant System The effectiveness of these improvements made to the new FC stack, specifically the improved MEA and reduced thermal mass, was verified in cold start tests of a 16-cell short stack from a temperature of -20 deg. C, -15 deg. C and -10 deg. C. The degree of improvement in cold start capability attained with the improved MEA was verified first. That was done by making a comparison of cold start performance between a 16-cell
short stack that combined the MEA of the 2005 model FC stack and carbon separator and a 16-cell short stack built with the MEA of the 2008 model FC stack and the 2005 model carbon separator. Then, to confirm the improvement in cold start capability due to the reduced thermal mass, a comparison of cold start performance was made with a 16-cell short stack that combined the MEA of the 2008 model FC stack and the metal separator.
The initial operating conditions can substantially affect the accuracy of the test results because the initial membrane water content before cooling down the stack has a large impact on cold start performance. Therefore, for these tests a sequence of experimental protocols was created to simulate the stack operation envisioned for ordinary vehicle use, and the same conditions were used in all the tests. Power generation was first performed at the vehicle's ordinary operating temperature and then stopped; the shutdown process was then carried out and the stack was cooled.
Because the performance of the 2008 model FC stack was improved at low humidification levels as shown in Fig. 13, the stack can actually be operated at lower cathode humidification than the 2005 model FC stack. However, in order to compare the cold start performance of both stacks under the same condition, the pre-cold startup operating conditions used were those of the 2005 model FC stack. Ultra-fine thermocouples were inserted in the 16-cell stack to confirm that, after cooling the environmental chamber to the cold start temperature (e.g. -20 deg. C), the stack interior was cooled to the temperature (e.g. -20 deg. C) before beginning the cold startup test.
For a short stack of around 16 cells, heat radiation from the end cells has a much more pronounced influence compared with an actual vehicle-mounted stack consisting of 400-500 cells. Therefore, a ceramic heater was installed for the end cells, and the heater output was adjusted so that the end cell temperature rise was approximately the same as that of the center cells. This made it possible to ignore the influence of the heat radiation characteristic of short stacks.
The experimental results obtained with the three 16-cell short stacks are shown in Fig. 46. In these tests, the load was drawn under a constant cell voltage, and a comparison was made of the time it took each stack to reach an output level equal to 50% of the rated power of a full stack. The results indicate that the improved MEA with a lower-EW membrane for the purpose of suppressing the ice formation rate in the cathode CL had the effect of shortening the startup time by 42%. The addition of the metal separator with a reduced thermal mass had the effect of shortening the startup time by an additional 40%. The effect of the reduced thermal mass works to improve cold startup capability by quickening the temperature rise of the stack. Furthermore, the resultant higher temperature also works to retard the ice formation rate in the cathode CL, as was noted earlier. The test results confirm that both effects combine to
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
substantially shorten the startup time more than what is attributable to the reduced thermal mass alone. The results for the 2008 model FC stack that incorporated both improvements show that the startup time was reduced by 65% compared with the 2005 model FC stack.
1.2 1
0.8
0.6 0.4 0.2 0
▲ 2005 Model MEA + Carbon Separator ■ 2008 Model MEA + Carbon Separator • 2008 Model MEA + Metal Separator
42% \
reduction \ \ \ \
40% »ч \ reduction^ чч \
■ А
i
was started from a temperature of -20 deg. C. The results indicate that power output of 40-50 kW was obtained in about 30-35 sec. from startup. This level of performance would pose virtually no problem for launching a vehicle in the real world.
60 50
Г
¥ 30
s
о
^ 20
10
0
20
25
35
40
-25 -20 -15 -10 -5
Startup Temperature (deg C)
Fig. 46. Time ratio to reach 50% of rated power from various subzero temperature for different MEAs and separator materials in short stack experiment [24]
These short stack experimental data validate the concept proposed earlier for improving cold start capability by increasing the product water uptake potential and uptake rate into the membrane and by reducing the cell thermal mass. The results verify that this concept is notably effective in obtaining higher power output faster during cold startup.
Having confirmed with the 16-cell short stack that the 2008 model FC stack provides improved cold start capability, a full-size stack was then incorporated into a prototype power plant system and subjected to cold start tests. The downsized water recovery device (WRD) described earlier was adopted for the system to provide a low level of humidification. Various measures were also taken to prevent the balance of plant (BOP) components from freezing, thereby making it possible to conduct cold start tests with this power plant system. Because thermocouples could not be inserted into the stack interior, a coolant containing a 50% ethylene glycol solution was circulated through the stack. It was confirmed that the stack interior was cooled to -20 deg. C before the test was started. The tests were conducted with the power plant system installed in a somewhat large cold room. The air supplied to the stack by a compressor was the room's ambient air at a temperature of -20 deg. C. In addition, the external air supplied to the cold room was first dehumidified and cooled to the same temperature as the room. This enabled the tests to be conducted under the same conditions as a vehicle placed in a -20 deg. C environment.
Fig. 47 shows the relationship between the startup time and power output of the power plant system incorporating the 2008 model FC stack when the system
30
Time (sec)
Fig. 47. Cold start experimental data from -20 deg. C for FC power plant equipped with 2008 model FC stack [24]
The prototype power plant system incorporating the 2008 model FC stack was then mounted in the X-TRAIL FCV shown in Fig. 48. The cold startup capability and operating performance of this improved X-TRAIL FCV were verified in subzero startup tests conducted on an environmental dynamometer and in cold-weather driving tests carried out at Nissan's Hokkaido Proving Ground. This improved FCV can be started at subzero temperatures as low as -20 deg. C, thanks to the 2008 model FC stack and some newly developed elements of the overall system.
Fig. 48. X-TRAIL FCV equipped with 2008 model FC stack [24]
Futre Work
Further downsizing of the FC stack and reduction of Pt loading are among the measures needed to lower the cost of FCVs. Toward that end, it will be necessary to increase the current density and to reduce the amount of Pt loading per unit area. Increasing the current density means raising it to a higher level for obtaining the same power output. For example, if the active area is reduced by half, the current density must be doubled to obtain the same level of power. The same is true for cold starts. As
noted earlier, raising the current density during cold startup affects the water balance inside the membrane by increasing the proportion of EOD and the water generation rate per unit time in relation to back diffusion. That, in turn, increases the ice formation rate in CL pores, causing cold start capability to deteriorate. Furthermore, Thompson et al. have reported that cumulative product water decreases when Pt loading per unit area is reduced [39]. These observations indicate that cost reduction of the stack conflicts with the assurance of cold start capability.
To address this trade-off, continued efforts will be made to improve the water uptake potential and uptake rate into the membrane and to reduce the cell thermal mass, as noted earlier. However, it would appear that there is a limit to how much the repercussions on performance due to increased current density and reduced Pt loading per unit area can be overcome. Along with reducing the cost in the future, it will also be necessary to consider FC stack startup from -30 deg. C or lower temperatures in order to popularize FCVs over wider geographical areas. That will make it necessary to improve the cold start capability of other components besides the membrane.
One key to resolving this issue is to elucidate the mechanism of ice formation in CL pores, which has a profound impact on cold start capability. Thompson et al. and Li et al. used a cyro scanning electron microscope (SEM) to observe the formation of ice in the CL in an effort to explicate the formation mechanism [39-41]. However, at the present time, the experimental results and interpretations in the literature are split between whether ice formation concentrates more toward the micro-porous layer (MPL) or more toward the membrane when the current density is increased. In this regard, various discussions have taken place based on the use of numerical simulation models, but the mechanism of ice formation has yet to be elucidated. We are proceeding with our own activities to use a cryo SEM to visualize ice formation in the CL under isothermal cold start conditions. Fig. 49 shows an example of a cyro SEM image. Ice formation on the MPL side is observed under the conditions used in this example. As indicated here, it is necessary to have a more detailed understanding of the physical phenomena involved in ice formation in the CL layer under certain given conditions.
In order to elucidate this mechanism, it will be necessary to advance the level of numerical simulation technology, including the treatment of interface phenomena, and to couple experimental and computational techniques for understanding ice formation better. That will require accurate knowledge of subzero physical properties, especially those of polymers. Concurrent with our cryco SEM observations, we are striving to obtain highly accurate data concerning the subzero properties of the membrane, including its a-X characteristic, water diffusivity, EOD coefficient and a-c characteristic.
We plan to proceed with our research activities in these areas in order to elucidate the mechanism of ice formation in the CL. Our aim for the future is to reconcile a lower FC stack cost with the assurance of cold start performance and also to secure cold start capability at even lower temperatures.
(4) Cost
Various phenomena occurring in the FC stack in relation to performance decay, power density and subzero start-up capability, which have been among the principal issues of FCVs, have been elucidated and measures for addressing them have been made clear. As a result, FC stack performance in these areas is approaching a level that is satisfactory for real-world application. At the present time, cost reduction is the major remaining issue.
Fig. 50 and 51 present examples of a cost breakdown for the FC powerplant system and the FC stack, respectively, based on cost analyses done by Sinha et al. [42]. The data show that the stack accounts for 50% of the overall FC system cost and that the electrodes account for 54% of the total stack cost. This indicates that the electrode cost represents approximately 25% of the overall FC system cost. One reason for the high cost of electrode materials can be attributed to the use of Pt catalysts. It is clear that the Pt loading must be reduced in order to lower the total cost.
Fig. 49. Cryo SEM image of cathode electrode cross section after cold start operation at -20 deg. C with 100 mA/cm2 [24]
Fig. 50. Example of FC powerplant system cost breakout [42]
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
12000
£10000
current distribution in the carbon
).7V, RH 40%
4 6
CCL thickness / um
10
Fig. 53 shows the change in reactant gas transport resistance at the cathode and anode electrodes as a function of the Pt loading [54]. These experimentally obtained data indicate that reactant gas transport resistance increased as the catalyst loading was reduced. This result implies that reducing the Pt loading decreases the effective area of the Pt catalyst, thereby increasing reactant gas transport resistance, especially that of oxygen. In future work, we plan to analyze the reaction and physical transport mechanisms that take place at the electrode catalyst layer with the aim of devising ways to reduce the Pt loading further while maintaining the desired performance and durability.
Fig. 51. Example of FC stack cost breakout [42]
The 2008 model FC stack halves the Pt loading compared with the previous stack [27]. That was accomplished by reducing the Pt loading of the anode electrode, which does not contribute much to stack performance, along with reducing the Pt loading of the cathode electrode.
However, since a large amount of platinum is still used per FCV, further efforts are needed to research and develop technologies for reducing the Pt loading to a lower level in the future
One approach to reducing the Pt loading further is to make more effective use of the platinum in the electrode catalyst layer. Fig. 52 shows the calculated current density distribution in the thickness direction of the cathode catalyst layer (CCL) under a low humidification condition, which is an effective way of reducing the FC system cost [43]. The results indicate that resistance to proton transport though the ionomer in the catalyst layer increases with increasing thickness, thereby localizing the reaction toward the electrolyte membrane. It is also seen that the current density in the Pt catalyst carbon support is distributed toward the outer surface. It is inferred from these results that only a portion of the platinum is being used effectively for its catalytic function, which suggests a possibility for reducing the Pt loading by making the reaction distribution more uniform.
0.35
0.2
Pt-loading mg cm'
0 12
■7
0.05
Fig. 52. Current density distribution in the cathode catalyst layer thickness direction [43]
Fig. 53. Reaction gas transport resistance in the catalyst layer [44]
Summary
(1) Along with electric vehicles (EVs), fuel cell vehicles (FCVs) will surely be a key solution for building a sustainable society with zero-emission mobility.
(2) There are four major issues that must be addressed in order to put FCVs on the market: (1) prevention of performance decay, (2) increase in power density, (3) improvement of subzero start-up capability and (4) reduction of costs.
(3) With regard to the issue of preventing performance decay, a method was established for estimating FC stack performance degradation, which provides a way of accurately estimating the durability of an in-vehicle stack. The main causes of stack performance decay were found to be in the MEA catalyst layer and include carbon support corrosion during start-up and Pt catalyst dissolution during load cycling and idling.
(4) With regard to increasing power density, a 2.0 kW/L FC stack has been achieved through performance enhancements obtained by adopting metal separators and by improving the MEA.
(5) As for improving subzero start-up capability, the physical phenomena taking place in a fuel cell during subzero start-up were analyzed and measures for improving subfreezing start-up performance were implemented. Test results confirmed that an FC system
incorporating the improved 2008 model FC stack can produce 40-50 kW of power within a start-up time of approximately 30-35 s.
(6) The major issue remaining to be addressed is cost reduction, for which reducing the amount of platinum catalysts used would be effective, as the electrode catalysts account for approximately 25% of the overall FC system cost. One way of reducing the Pt loading is to use Pt more effectively. Toward that end, it will be important to further analyze the reaction and physical transport mechanisms taking place at the electrode catalyst layer.
Acknowledgements
The elucidation of the cold start mechanisms of the MEA described in this paper resulted from joint research work conducted by Nissan Motor Co., Ltd. and The Pennsylvania State University. The authors would like to thank Dr. Chao-Yang Wang of The Pennsylvania State University for his excellent guidance and all the research team members for their significant contributions.
References
1. Arai T., Sakai H., Ikezoe K., Aoyama T., Yanagisawa M., Kanesaka H. Current status and Future issues of fuel cell vehicle development at Nissan, Fisita, F2010-A-070, 2010.
2. Aoyama T., Iiyama A., Shinohara K., Kamegaya S., Yamamoto S. and Ban Y. (2008). Status of FCV Development at Nissan and Future Work. SAE 200801-0423.
3. Shimoi R., Aoyama T., Iiyama A. (2009). Development of Fuel Cell Durability based on Actual Vehicle Test Data: Current Status and Future Work. SAE 2009-01-1014.
4. Iiyama A. (2007). FC Development in Nissan, Proceedings of the FC Expo Technical Conference, Tokyo, February 2007.
5. Reiser C.A., Bregoli L., Patterson T.W., Yi J.S., Yang J.D., Perry M.L., Jarvi T.D. (2005). A Reverse-Current Decay Mechanism for Fuel Cells. Electrochem. Solid-State Lett., 8 A273-A276.
6. Meyers J.P., Darling R.M. (2006). Model of Carbon Corrosion in PEM Fuel Cells. J. Electrochem. Soc., 153 A1432-A1442.
7. Darling R.M., Meyers J.P. (2003). Kinetic Model of Platinum Dissolution in PEMFCs. J. Electrochem. Soc., 150 A1523-A1527.
8. Darling R.M., Meyers J.P. (2005). Mathematical Model of Platinum Movement in PEM Fuel Cells. J. Electrochem. Soc., 152 A242-A247.
9. Wang X., Kumar R., Myers D.J. (2006). Effect of Voltage on Platinum Dissolution. Electrochem. SolidState Lett., 9 A225-A227.
10. Yasuda K., Taniguchi A., Akita T., Ioroi T., Siroma Z. (2006). Platinum dissolution and deposition in the polymer electrolyte membrane of a PEM fuel cell as studied by potential cycling. Phys. Chem. Chem. Phys.,
8 746-752.
11. Borup R.L., Davey J.R., Garzon F.H., Wood D.L., Inbody M.A. (2006). PEM fuel cell electrocatalyst durability measurements. J. Power Sources, 163 76-81.
12. Mitsushima S., Kawahara S., Ota K., Kamiya N.
(2007). Consumption Rate of Pt under Potential Cycling. J. Electrochem. Soc., 154 B153-B158.
13. Uchimura M., Kocha S. (2007). The Impact of Cycle Profile on PEMFC Durability. ECS Trans., 11 (1) 1215-1226.
14. Ferreira P.J., la O' G.J., Shao-Horn Y., Morgan D., Makharia R., Kocha S., and Gasteiger H.A. (2005). Instability of Pt/C Electrocatalysts in Proton Exchange Membrane Fuel Cell. J. Electrochem. Soc., 152 A2256-A2271.
15. Ohma A., Suga S., Yamamoto S., and Shinohara K. (2007). Membrane Degradation Behavior during Open-Circuit Voltage Hold Test. J. Electrochem. Soc., 154 B757-B760.
16. Ohma A., Yamamoto S., Shinohara K. (2007). Analysis of Membrane Degradation Behavior During OCV Hold Test. ECS Trans., 11 (1) 1181-1192.
17. Ohma A., Suga S., Yamamoto S., Shinohara K. (2006). Phenomenon Analysis of PEFC for Automotive Use (1): Membrane Degradation Behavior During OCV Hold Test. ECS Trans., 3 (1) 519-529.
18. Ohma A., Yamamoto S., and Shinohara K.
(2008). Membrane degradation mechanism during open-circuit voltage hold test. J. Power Sources, 182 39-47.
19. Endoh E., Terazono S., Widjaja H., Takimoto Y. (2004). Degradation Study of MEA for PEMFCs under Low Humidity Conditions. Electrochem. Solid-State Lett., 7 A209-A211.
20. Teranishi K., Kawata K., Tsushima S., Hirai S. (2006). Degradation Mechanism of PEMFC under Open Circuit Operation. Electrochem. Solid-State Lett., 9 A475-A477.
21. Inaba M., Kinumoto T., Kiriake M., Umebayashi R., Tasaka A., Ogumi Z. (2006). Gas crossover and membrane degradation in polymer electrolyte fuel cells. Electrochim. Acta, 51 5746-5753.
22. Mittal V.O., Kunz H.R., Fenton J.M. (2006). Is H2O2 Involved in the Membrane Degradation Mechanism in PEMFC? Electrochem. Solid-State Lett.,
9 A299-A302.
23. Sugawara S., Maruyama T., Nagahara Y., Kocha S.S., Shinohara K., Tsujita K., Mitsushima S., Ota K. Performance decay of proton-exchange membrane fuel cells under open circuit conditions induced by membrane decomposition. J. Power Sources, in press.
24. Ikezoe K., Tabuchi Y., Kagami F., and Nishimura H. Development of an FCV with a New FC Stack for Improved Cold Start Capability. SAE 2010-01-1093, 2010.
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
25. Yaginuma M., Miyazawa A., Tada E., and Inoue H. Influences of Gas Diffusion Layer on Corrosion of Metallic Bipoler Plate. The 56th Japan Conference on Materials and Environments, Abstr., B-215, 199, 2009.
26. Ito M., Mure T., Nishikata A., and Tsuru T., Yaginuma M. and Miyazawa A. Corrosion potential measurement of stainless bipolar plate in PEFC environment. The Electrochem. Soc. Jpn., Abstr., 1G24, 115, 2009.
27. Miyazawa A., Ikezoe K., Okuyama Y., and Yanagisawa M. The Development of High Power Density and Low Cost New Fuel Cell Stack. JSAE Annual Congress (Spring), Proceedings No. 22-09, 10920095161, May 20, 2009.
28. Miyazawa A., Yaginuma M., and Tada E. Sensitivity evaluation of metallic ions to degradation on cell performance. The 50th Battery Symposium in Jpn., Abstr., 1G05, 463, 2009.
29. Yaginuma M., Himeno T., Miyazawa A., and Arai T. Sensitivity evaluation of metallic ions to degradation on cell performance. The 49th Battery Symposium in Jpn., Abstr., 1D22, 208, 2008.
30. Kagami F., Hishinuma Y. andChikahisa T. (2002) Therm. Sci. Eng., 10, 25.
31. Oszcipok M., Riemann D., Kronenwett U., Kreideweis M. and Zedda M. (2005). Statistic analysis of operational influences on the cold start behavior of PEM fuel cells. J. Power Sources, 145, 407.
32. Mao L. and Wang C.Y. (2007). Analysis of Cold Start in Polymer Electrolyte Fuel Cells. J. Electrochem Soc. 154(2) B139-B146.
33. Tajiri K., Tabuchi Y. and Wang C.Y. (2007). Ishothermal Cold Start of Polymer Electrolyte Fuel Cells. J. Electrochem Soc. 154(2) B147-B152.
34. Mao L., Wang C.Y. and Tabuchi Y. (2007). A Multiphase Model for Cold Start of Polymer Electrolyte Fuel Cells. J. Electrochem Soc. 154(3) B341-B354.
35. Tajiri K., Tabuchi Y., Kagami F., Takahashi S., Yoshizawa K. and Wang C.Y. (2007). Effect of operating and design parameters on PEFC cold start. J. Power Sources, 165 279-286.
36. Zawodzinski T.A., Springer T.E., Davey J., Jestel R., Lopez C., Valerio J., and Gottesfeld S. (1993). A Comparative Study of Water Uptake by and Transport through Ionomeric Fuel Cell Membranes. J. Electrochem. Soc., 140, 1981.
37. Motupally S., Becker A.J., and Weidner J.W. (2000). Diffusion of Water in Nafion 115 Membranes. J. Electrochem. Soc., 147, 3171.
38. Tabuchi Y., Takaichi S. and Kubo N. (2008). Limiting Current Density Under Low Humidity Condition, Meet. Abstr. Electrochem. Soc. 802 793
39. Thompson E.L., Jorne J., Gu W. and Gasteiger H.A. (2008). PEM Fuel Cell Operation at -20C. I. Electrode and Membrane Water (Charge) Storage. J. Electrochem Soc. 155(6) B625-B634.
40. Thompson E.L., Jorne J., Gu W. and Gasteiger H.A. (2008). PEM Fuel Cell Operation at -20C. II. Ice Formation Dynamics, Current Distribution, and Voltage Losses. J. Electrochem Soc. 155(9) B887-B896.
41. Li J., Lee S. and Roberts J. Electrochim Acta, (2008). Vol. 53. P. 5391.
42. Sinha J., Lasher S., and Yang Y. Direct Hydrogen PEMFC Manufacturing Cost Estimation for Automotive Applications. DOE Hydrogen Program Review/USA, Project ID #FC_31_Sinha, May 21, 2009
43. Ohma A., Mashio T., Ono Y., Sato K., Iden H., Sakai K., Kanesaka H., and Shinohara K. Estimation of the Pt effectiveness factor in a cathode catalyst layer for PEMFC. The 50th Battery Symposium in Japan, 2F03, Nov. 30, 2009.
44. Sakai K., Sato K., Mashio T., Ohma A., Yamaguchi K., and Shinohara K. Analysis of Reactant Gas Transport in Catalyst Layers: Effect of Pt-loadings. ECS Trans. 25 (1), 1193, 2009, Reproduced by permission of The Electrochemical Society.
Contents
Global Warming and CO2 Reduction FCV Progress and Publicity Activities Durability and Cost Reduction Future Perspectives and Summary
2
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Global Environment Issues
Environment issues have been expanding from regional issues to global issues.
★1988 Montreal Protocol ( Ozone layer protection )
★2005 Kyoto Protocol ( CO2 reduction )
Low 1970
Regional Issues
1980
1990
2000
Present
Global Issues
Reference : John Elkington Presetation (Sustainability)
3
Effects of Global Warming
■ Social Life will be affected seriously
z:-г
с
+4
+3
+2
I
Malaria : 50 - 80 mil. people
20% increase of rainfall : Flood damage increase. Greenland ice sheet ! Almost melt, 7m increase of Sea level
Coral reef ! Bleaching in wide area > Agriculture ! Decrease of production in middle latitude area
Sea level: +10-45cm (increase), Rainfall : A5-10% (decrease) Flood risk area 23-29% increase in Bangladesh lowland.
=4 Source) IPCC 3rd Report
Международный научный журнал «Альтернативная энергетика и экология» № 9 (89) 2010 © Научно-технический центр «TATA», 2010
Potential Global Warming Scenario
Theory (from IPCC 3rd report)
Average global temperature will increase up to +2°C on the basis of BAU*
I
Atmospheric CO2 concentration must be stabilized under 550 ppm level (450ppm in 4th report)
^ Temp. Increase
^^^^ ¡ft
Source) IPCC 3rd and 4th Assessment Report
*BAU (Business as Usual): Scenario in which global warming continues without countermeasures being taken
90 £ 80
■2 70
(0
I 60
SS*
8^40
« 30
o 20 O
10
2000 2050 2100 2150 2200 2250 2300 CO2 Emissions
BAU
Reduce to 1/5 by 2100
2000 2050
5
Long Term Goal for Reducing CO2
To reduce C02 emissions from all new vehicles by 90%,
• Short & mid term : Internal Combustion Engine (ICE) efficiency
• Long term : Electric Powertrains
C02 free energy (Collaboration with other sectors)
1001
80
u a> 3 a
5
60
40
20
0
2000 2010 2020 2030 2040 2050
100
80 60 40 20 0
o o
3
■a a m
es
O o
7D <d
-C
5
o
0 5
ICE
HEV
EV
FCV
6
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Structure of FCV
■ Electricity is generated at FC stack by the electro-chemical reaction of oxygen and hydrogen, the electricity drives a motor.
Expectation of FCV
■ Zero Emission Vehicle
■ Various Energy Source can be Applied
■ Good Energy Efficiency
■ Low Noise and Vibration
■ Short Time Refueling, Long Range
8
Zero Emission Vehicle
■ Both EV/FCV(BEV/FCEV) Create Zero Emission
9
Contents
FCV Progress and Publicity Activities
10
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
FCV Development Activities
I Next generation FCV with improved in-house developed stack (beginning of next decade, starting from US and Japan markets)
Range :
160 km
Acceleration* : 25sec.
200 km
20sec.
18sec. Almost the same level as gasoline time from 0km/h to 100km/h powered vehicle
mance
X-TRAIL FCV FY2005 Model
11
X-trail FCV 05 Model
■ Performance level similar to ICE vehicles
■ Operating over 3 years in the US and Japan (demonstration program, etc.)
■ Limited leasing to customers
Major Performance
Max. Speed : 150km/h
Cruising Range: 370km(@35MPa)4 500 km(@70MPa)
12
Международный научный журнал «Альтернативная энергетика и экология» № 9 (89) 2010 © Научно-технический центр «TATA», 2010
Demonstration Program
■ Demonstration is being continued in the world.
■ Total mileage is over 900,000km. (July, 2010)
Nissan FCV @CaFCP (California Fuel Cell Partnership )
13
Demonstration Program (Leasing Car)
Nikko City
Yokohama City
Sacramento Coca-Cola Bottling Co.,Inc
Kanagawa-Toshi-Kotsu ( First FCV Limousine in —business in the world)
14
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Joint FCV Development with Renault
■Developed Renault FCV (Scenic based) jointly. ■From June, 2008 under demonstration in EU.
1Б
New FC Stack (August, 2008)
iFeatures of technology
> Doubled power density (metal separator, modified membrane)
> Low cost (Half Pt loading, etc.)
Specification
Max Power
Volume
Mass
130kW
68L
86kg
New FC Stack [Power Density]
X1.7
[Cost]
X2
/
/
✓
35% ч
35%
03 Model 05 Model 08 Model
03 Model 05 Model 08 Model
-1б
FC Stack Development Activities
■Power density is top level
3.0
d 2.5
£
> 20
W
H 15
№
§ 10 Q.
0.5 0
Nissan
Honda / ✓ ✓ /
A 1 D8model
05mod<
2002 2004 2006 2008 2010 2012 2014
FY
Vehicle test with the New FC stack
■ Cold Weather Test is executed by using the vehicle equipped with the New FC stack .
7/
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
17
18
FCV Publicity Activities
2009 Yokohama Women's Marathon US Hydrogen Road Tour 09
19
Contents
Durability and Cost Reduction
20
Key Factor of FC Stack Performance Degradation
Operating mode Degradation Main causes
Start-up • Cathode carbon support corrosion • Cathode potential elevation by anode gas replacement
Load cycling • Cathode catalyst (Pt) dissolution • Cathode potential cycling
Idling (low current) • Membrane electrolyte decomposition •H2O2 formation by gas crossover
The key factor of FC stack performance degradation is platinum dissolution and carbon corrosion
Shimoi, et al. ; J.S.A.E 2009 Spring-meeting. 107-20095032
21
Estimation of Stack Degradation
Comparison between observed FCV degradation and estimated degradation.
Ratio of Operation mode on degradation. (Calculation based on gasoline engine vehicle data on normal public road)
Load cycling (Estimation)
□ Idling (Estimation)
□ Start/stop cycles (Estimation) OVehicle data_
o
XT
Japan#1
Japan#2
il
*Data from FCVs
with detailed
operation record and
more than 1.5 years
operation 1%
Degradation (Estimation)
Degradation of on-road vehicles can be very well estimated.
Source: JSAE Spring Meeting, April, 2009
22
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Carbon Corrosion Mechanism
, ,-e—^ Anode
1 H2^> H2^2H++2e- H2 " O2 O2+4H++4e-^2H2O
' К : ÎH+ Anode catalyst layer Membrane
1 O2+4H++4e-*2H2O oj 2H2O*<V4H++4e 2 2 2, C+2H2O^CO2+4H*+4e- Cathode catalyst layer
1 ^ e- , Cathode
Electromotive is generated
When the FC stack start-up, the carbon corrodes because a part of the cathode becomes high potential.
23 Shimoi, et al. ; J.S.A.E 2009 Spring-meeting. 107-20095032
Pt Dissolution Mechanism
Pt in the cathode catalyst layer dissolves by a potential cycle at the load cycle and high potential in idle-stop.
Shimoi, et al. ; J.S.A.E 2009 Spring-meeting. 107-20095032
24
FCV Cost Reduction
T EV related parts cost reduction
( System simplification/spec down
FCV unique parts cost reduction)
~25
FC System Cost Reduction
Balance of Plant 19%
Fuel Management 6%
Thermal Managemen 7%
Water Management 4%
Air Management 10%
Membrane 4%
Balance of Stack 5%
• Total $ 8, 000/ 80kw
• Catalyst share 30%
• Necessary to make Pt 1/10 (10g/vehicle) due to Pt resource limitation.
* Stack component (MEA^ separator, etc.)
• System component cost reduction. (Material development, simplification)
FC System Cost estimation : D0E(2008)
26
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
System Simplification/Size Reduction
* Higher temperature
& lower/non humidification operation
^ Water management
(Membrane, GDL, Catalyst layer,
retain and exhaust capability)
* Contact resistance
^ Separator/GDL/Catalyst layer
contact resistance reduction
* Enhance catalyst activity
^ Specific activities, Pt utilization
Fuel Cell evolution on material and structure is necessary
27
Improvement of Cell Performance
gas channel
gas channel
mol density
Resistance of Water Electroosmotic flux
Resistance of Water ______________
water transfer concentration water transfer concentration gradient gradient
Improvement of water transfer from anode side to cathode side due to back diffusion in membrane -►Higher relative humidity in anode
side enable operation at higher current density
Miyazaki, et al. ; the 50th Battery Symposium in Japan
1.0 1.5 2.0 г. 5 current density [A/cm2]
I 2-Û e
* o.o
fr a = - 0.135 LJ Sample A
corresponds to about 30 % of water liquid о Sample В
Д
o.o
0.5
1.0 1.6 2.0 current density [A/cm2]
3.0
28
Международный научный журнал «Альтернативная энергетика и экология» № 9 (89) 2010 © Научно-технический центр «TATA», 2010
Cost Reduction with Stack Performance Improvement
■ Improve power generation and enable low humidifying operation
■ Achiev miniaturization of humidifier and reduce cost
Cathode Inlet RH (%%
Miyazawa, et al. ; J.S.A.E 2009 Spring-meeting. 109-20095161
29
Catalyst Utilization Improvement
Current by Cathode reaction (¡orr) > Function of effective Catalyst surface, O2 ,
lom
concentration, and Proton concentration Definition of Pt utilization considering reaction distribution
( C s ^ O 2 ( C H + \
C ref O 2 V 2 / C ref
v /
exp
' anF x
RT
Pt Utilization Sum of actual current (reaction) distribution Z i
(ReaS speed Sum of ideal (constant concentration = bulk condition) current Z i
actual
ideal
Concentration and current distribution in CL
Pt loaded carbon
Membrane
Cathode CL
Cathode GDL
bul k
t distribution ant reaction ntration
bul k
Effect of low mass transport (low concentration) is very large at catalyst layer(random network structure)
Z ii
ideal
Ideal constant concentration zu -rent distribution iactual Actual current distributio Primary Pore
10nm) Lack of mass transport media(ionomer, water
o
50 nm
lonomer covering or water state affect a lot for current distribution
Low concentration . , of reaction material '' (proton)
lonomer -
| lower the generated current
n
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Opportunity for Further Reduce Pt in MEA
ln-situ Pt utilization could be very low under dry and high current density conditions
100 90 ^ 80
5 70 I 60
3 50 * 40 ъ 30 20 10 0
RH90% RH40%
0.90 0.70
iR- free Cell Voltage I V
12000
|ioooo
RH40%0.7V
current distribution in the carbon
0 2 4 6 8 * Mem. CCL thickness I um
3l
Contents
Future Perspectives and Summary
32
Future Perspectives
Durability technology will become feasible around 2010. ■ Cost reduction will be the challenges towards 2015.
Technology Introduction/Expansion Penetration Development Phase Phase Phase (2010s) (2020s) Practicality (Driving performance^ Cruising Range)
Vehicle Long durability
Issues 2015 FC cost drastically reduction Innovative Hydrogen Storage
Infrastructure Hydrogen cost reduction . , , Infrastructure Issues expansion 33
FCCJ Scenario (Released at JHFC Seminer 2009)
Commercialization Scenario for FCVs and H2 Stations
Phase 1
Technology Demonstration [JHFC-2]
2010
Phase 2
Technology & Market Demonstration [Post JHFC] 2011 2015
•Solving technical issues and promotion of review regulations (Verifying & reviewing development progress as needed)
•Verifying utility of FCVs and H2 stations from socio-economic viewpoint
Phase 3
Early Commercialization
[Starting Period] 2016
[Expansion Period] 2025
•Expanding production and sales of FCVs while maintaining convenience of FCV users
•Reducing costs for H2 stations and hydrogen fuel
•Continuously conducting technology development and review of regulations
Phase 4
Full Commercialization
[Profitable business Period 2026
Contribute to diversity of energy sources and reduction of CO2 emissions
I nc rea senu m bersof FCV and H2_slalioris based on profitable business
Costs for H2 station construction and hydrogen reach targets, making the station business viable. (FCV 2,000 units/station)
Period in which preceded H2 station building is necessary
Increase of FCV numbers through introduction of more vehicle models
Note: Vertical axis indicates the relative scale between vehicle number & station number.
Year
Precondition: Benefit for FCV users (price/convenience etc.) are secured, and FCVs are widely and smoothly deployed
-34
FCCJ: Fuel Cell Commercialization Conference of Japan
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Summary
■ FCV technology shows steady improvement.
■ Low humidification operation and higher durability could be achieved by mechanism understanding of water transport in MEA and of degradation of carbon support and Pt.
■ Still lots of researches needed to overcome technical barriers for cost reduction.
■ Also it is necessary to improve FCV publicity and to promote infrastructure development.
35
Nissan FCV Spec Sheet
R'nessa FCV
Model
N/A
Dimensions (L x W x H)
4680 x 1765 x 1625 mm
Curb Weight
2700kg
Seating Capacity
2 Persons
Traction Motor Type
Neodymium permanent magnet synchronous motor
Max. Power
62kW
Max. Torque
160N-m
Fueling System
Methanol
Max. Speed
100Km/h
Cruising Range
N/A
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Model
N/A
Dimensions (L x W x H)
N/A
Curb Weight
N/A
Seating Capacity
5 Persons
Traction Motor Type
Permanent magnet synchronous motor
Max. Power
75kW
Max. Torque
N/A
Fueling System
Compressed hydrogen gas
Max. Speed
120 Km/h
Cruising Range
Over 200 km
Model N/A
Dimensions (L x W x H) 4465 x 1765 x 1790 mm
Curb Weight N/A
Seating Capacity 5 Persons
Traction Motor Type Coaxial motor integrated with reduction gear
Max. Power 58 kW
Max. Torque N/A
Fueling System Compressed hydrogen gas 35 MPa
Max. Speed 125 Km/h
Cruising Range Over 200km
International Scientific Journal for Alternative Energy and Ecology № 9 (89) 2010
© Scientific Technical Centre «TATA», 2010
Model
FT30
Dimensions (L x W x H)
4485x1770x1800 mm
Curb Weight
1950 kg
Seating Capacity
5 Persons
Traction Motor Type
Coaxial motor integrated with reduction gear
Max. Power
85 kW
Max. Torque
280 N-m
Fueling System
Compressed hydrogen gas 35 MPa
Max. Speed
145 km/h
Cruising Range
Over 350 km
Dimensions (L x W x H)
4485 x 1770 x 1745 mm
Traction Motor Type
Coaxial motor integrated with reduction gear
Curb Weight
1790 (1860) kg
Seating Capacity
5 Persons
Max. Power
90 kW
Max. Torque
280 N-m
Fueling System
Compressed hydrogen gas 35 MPa (70 MPa)
Max. Speed
150 km/h
Cruising Range
370 km (500 km)
Specs of 70 MPa vehicle are in parentheses
Model
FAT30
fXD — TATA — CXJ
