Further, electric power trains, principally allow for recuperation of braking energy ... the electrical side, the electric machine is connected to the fuel cell and super.
Fuel Cell Supercap Hybrid Electric Power Train Felix N. Büchi*, Akinori Tsukada*, Paul Rodatz**,Olivier Garcia**, Martin Ruge**, Rüdiger Kötz*, Martin Bärtschi,* Philipp Dietrich*, * Paul Scherrer Institut, CH-5232 Villigen PSI Switzerland ** Swiss Federal Institute of Technology, CH-8092 Zurich, Switzerland
Abstract Fuel cells have the potential to change the propulsion system for cars. In a joint project Paul Scherrer Institut (PSI), ETH Zürich, FEV Motorentechnik, and Volkswagen have developed the fuel cell hybrid vehicle “Hy.Power”. Hy.Power is a technology platform for the demonstration of a powertrain combining a hydrogen fuel cell system and a double-layer capacitor arrangement. The characteristics of this powertrain, is presented and analyzed.
1 Introduction In the last decade concern about emission of gases contributing to the global warming has increased (Kyoto protocol). Transportation produces a significant share of the emissions. These emissions are related to the efficiency of the power train and the fuel used Due to high part load efficiency fuel cell systems can contribute to improving the efficiency of the power train of a passenger car, which is often operated in part load. Further, electric power trains, principally allow for recuperation of braking energy, which can offer an additional increase of efficiency in the range of 1015% in city-cycles. However with pure fuel cell systems no recuperation is possible. Besides increasing efficiency, the use of an electric storage system allows to reduce the power of the fuel cell system, because the stored energy can be used for acceleration. Further recuperation is a potential step to increase mileage of electric passenger cars. The concept of a hybrid-electric powertrain, which consists of a fuel cell and an electrical energy storage device (Fig. 1), opens the possibility to design the performance level of the storage device and the fuel cell system device for different purposes. This additional design flexibility can be used to optimise costs by substituting the more expensive device with the cheaper one. For the
1
electric storage device a system with high power capability and high efficiency is needed. A super capacitor system can offer these properties. The research powertrain presented in this paper, which has been realised as a technology platform is used to explore the performance of new materials and system architectures to give insights for further development.
2 Powertrain The powertrain lay out is shown in Figure 1. The car is driven by an electric motor which is connected to the wheel by a constant reduction transmission. On the electrical side, the electric machine is connected to the fuel cell and super capacitor systems. Because the voltage of the fuel cell system and the super capacitor are different, they need to be connected by means of a power electronic device to adapt the voltages. To allow recuperation of the braking energy in the super capacitor the electric machine needs at least a 2-quadrant configuration. With the chosen 4-quadrant configuration the transmission can be simplified and the reverse gear can be substituted by the electric machine. Hydrogen is used as the fuel, therefore a tank system is needed within the vehicle but no reformer unit is required. The disadvantage of the direct hydrogen storage in the vehicle is the low energy density of this fuel, at least in it’s gaseous state.
Figure 1: Schematic view of the PE fuel cell and super capacitor hybrid powertrain.
2
3 Fuel Cell System The fuel cell system can be divided in the following four subsystems: • stack: electro-chemical energy converter unit. • air-subsystem: supply of the process air at the needed pressure, temperature and humidity. • H2-subsystem: supply with H2 fuel at required pressure and flowrate. • cooling-subsystem: cool all other subsystems adequately. • control-subsystem: dynamic control of all active elements. 3.1
Stack development
The development of stack-technology can be divided in two areas : (i) Electrochemical components and development of the preparation procedures for these parts. Beside the high performance, minimum effort for the preparation is important; (ii) Development of a new bipolar plate (BIP) in order to build volume and weight efficient stacks. Again, beside the performance as a relevant criterion, attention needs to be paid also on the optimization of the manufacturing process of the bipolar plates. Adequate performance under the conditions of the mobile fuel cell system, low degradation potential over time and the possibility for optimized preparation procedures have guided the selection process of the electrochemical components. Commercially available membranes (Nafion® 112, DuPont) and electrodes (ELAT, E-Tek) were evaluated and the respective preparation and assembly procedures developed. At the conditions of the mobile fuel cell system with respect to process gas humidification a specific power density of 320 mW/cm2 (64 W/cell) was obtained at different pressure levels (Fig. 2). The BIP is a multifunctional part, which represents the biggest volumetric part of the stack. The BIP has to distribute the air and hydrogen to the membraneelectrode-assembly (MEA), support the cooling of the MEA, avoid the mixing of the different media or leaking to the exterior and conduct the current between electrochemical cells. Minimal material requirements for such BIP are: • •
Electric conductivity ≥ 10 S/cm Heat conductivity ≥ 20 W/m K
3
100 90 80 70 60 50 3.0 bar 2.5 bar 2.0 bar 1.5 bar
40 30
Cell Power [W]
Cell Voltage [V]
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
20 10 0
0
20
40
60
80
100
120
Current [A]
Figure 2: Current/voltage and current/power curves for a single cell at gas pressure variations. Gas stoichiometrics : λH2 = 2, λair = 2; gas dew points (DP): DPH2 = 50 °C, DPair = 60 °C; cell temperature 70 °C.
• •
Gas tightness: permeation < 10-5 mbar l/s cm2 Corrosion resistance in contact with an acidic electrolyte, oxygen, heat, and humidity.
The specified values for electric and heat conductivities are required to keep the voltage loss in the BIP below 3% at full load and ensure low temperature gradients, respectively. Optimization criteria for the design process are minimal volume (slim), minimal weight (light), optimal gas-supply (flow field) and minimal cost (Cheap material and short production cycle). Based on the science and technology developed for a single cell with a power of less than 100 W, a converter being the heart of a power train for a fuel cell electric car was developed and realized. Single cells are stacked in series through the optimized bipolar element [1, 2] to multi-kW stacks. The stacks then have been arrayed through thoughtful manifolding to a multi-10 kW system. The scale-up from a single cell (64 W) to a 125 cells stack (8 kW) was realized with small deviations between the stacks. Figure 3 shows the current/voltage characteristics of the 6 stacks integrated in the stack array (Fig. 4) and the comparison to a single cell under standard test conditions (stack temperature 70 °C, cooling water exit temperature ∆T 5-8 °C, gas pressures 2 barabs (exit), stoichiometrics of 2 for both process gases, and dew points of 55° and 50° C for hydrogen and air respectively). The variation between the different stacks is
4
fairly low (Fig. 3). The results from Figure 3 show that the design of bipolar plates and stacks as well as preparation procedures are well scalable.
1.0
Cell Voltage [V]
0.9 0.8 0.7 Stack 1 Stack 2 Stack 3 Stack 4 Stack 5 Stack 6 Single Cell
0.6 0.5 0.4 0.3 0
20
40
60
80
100
120
Current [A]
Figure 3: Current/voltage characteristics of a single cell and of the six stacks (averaged of the 125 cells) used for the stack array.
The modularity of a fuel cell system however, does also have disadvantages. The main disadvantage is the large number of parts, which are needed for powerful systems. The array of six stacks, shown in Fig. 3, contains more than 5000 parts. Most of these parts have to be handled, prepared, controlled and finally assembled individually. Except for few standard parts such as screws, springs and tie rods, all parts are individually designed. The six stacks were assembled to an array. In this array the stacks are connected gas wise in a parallel. Electrically they are connected as two parallel strings of 3 stacks in series in order to match the voltage requirements of the power train. For efficient manifolding of the process gases and the cooling liquid, requiring 6 connections to each stack, these media are connected to all stacks through a manifolding plate of a thickness of less than 10 cm, delivering gases and coolant liquid with equal pressure drop to all stacks. Fig. 4 shows the set-up of the 6-stack array including manifolding plate. The weight of the complete array is 185 kg (stacks 140 kg and manifolding plate and structure 45 kg).
5
Figure 4: Array of six stacks of 125 cells with manifolding plate (below and left side). Total weight 185 kg.
3.2
Hydrogen supply system
The hydrogen supply system has to provide enough fuel to the stacks under all operating conditions. Further the system has to remove water droplets from the anode side of the cells. Pure hydrogen is stored as compressed gas at a maximum pressure of 350 bar. The stacks are operated at nominal 2 bara and at temperatures up to 70°C. The hydrogen pressure is controlled as a function of the pressure at the air side. The simplest arrangement to supply the fuel cell with hydrogen is a dead-end system (arrangement A in Fig. 5). In this arrangement, only the hydrogen, which is needed to sustain the reaction is fed to the stacks. The dynamics are moderate. The dynamics can be enhanced by an excess flow of hydrogen through the stacks. To avoid the excess hydrogen being released to the surroundings it can be recirculated to the stack entrance by means of a fan or pump which compensates the pressure drop across the fuel cell. To eliminate the parasitic power of the fan the hydrogen can also be recirculated using an ejector.
6
A
Valve "a"
H2
Sensor 1
Magnetic valve M
P
PEFC
B Valve "a"
P
H2
PEFC Pump
C
Valve "b"
Vessel
M1
Sensor 2 P
M2
Vessel
Sensor 1
P
H2 Valve "a"
PEFC
Figure 5: Schematics of different hydrogen feed systems. A: standard system; B: system for low pressure pulses; C: system for high pressure pulses.
With purging (opening of valve M in arrangement A) not only inert gases are released from the system but shock waves are generated across the ventilation valve. The pressure difference between the sides of the valve induces a temporary flow through the manifold in the stack and stack array and the fuel cell flow field. Thereby, any water droplets that may have formed inside the fuel cell are dispersed. Further, the liquid water particles are blown out from the fuel cell, allowing for the delivery of additional hydrogen and thus preventing reactant starvation in parts of the cell. In addition to this the diffusion layer, which is situated between the flow channel and the MEA is dynamically inflected by the pressure wave. Arrangement B shows a system that is able to generate such pressure waves using a vacuum inside a vessel. A magnetic valve is installed between the fuel cell stack and the vacuum vessel. The pressure drop across this valve is similar to the pressure drop across the purging valve. Therefore the same effect as
7
purging to the environment is achieved. The parasitic power loss by the pump is 5 to 10 times lower than the fan used to recirculate hydrogen at constant pressure to maintain the same level of perofmance. If the vessel pressure is higher than the pressure in the stacks purge waves can be generated along arrangement C. The energy of the high pressure hydrogen storage device is transferred to the higher pressure vessel to support the waves above the stack pressure. For optimum performance, in the realized fuel cell system arrangements A, B, and C, together with an ejector are implemented.
3.3 Air supply system The air supply system is developed by FEV Motorentechnik GmbH and is schematically shown in Figure 6 and described in detail in [3]. For the air supply an Opcon screw compressor is employed. Water is injected at high pressure into the pressure side of the compressor to cool the hot air to cell temperature. As a side effect, the air is also humidified which is beneficial for the operation of the stacks. Water for injection is recovered from the exit air in a water separator.
Compressor
Pressure Control Walve
F
Water Separator
Water Injection Pump
Figure 6: Schematic lay-out of the air-loop.
3.4 Cooling system The fuel cell system is cooled with de-ionised water, which is circulated by a speed variable electrical pump. The heat transfer to the environment is realized by an air-water heat exchanger and by two separately controlled fans.
8
Figure 7 : Schematic of the complete fuel cell system, including hydrogen tanks.
4 Supercapacitors Super capacitors utilizing an organic electrolyte and activated carbon as electrode material [4] were developed for the present application in collaboration with montena components SA. The single capacitor cells had a minimum capacitance of 1500 F and a nominal voltage of 2.5 V. The internal equivalent resistance (ESR) was 1 mΩ typically. From the Ragone plot a maximum specific energy of 5.3 Wh/kg and a maximum specific power of 4.8 kW/kg was determined. Figure 8 shows the capacitor cell, which has a diameter of 50 mm and a length of 150 mm. Both connectors are arranged on one side.
9
Figure 8: 1500 F super capacitor cells used for the powertrain. In order to meet the max. voltage demand (360 V) of the powertrain, 141 capacitor-units were connected in series, corresponding to an average voltage across each capacitor unit of 2.55 V. In order to fulfill the energy demand each capacitor-unit consisted of two cells connected in parallel. As shown in Figure 9 the capacitors were assembled in two modules, one with 140 and the other with 142 capacitor cells. The two modules are connected in series and had a total weight of 168 kg made up of 110 kg capacitor cells, balancing electronic and electrical contacts and 58 kg for the metal housing, contactors, fuses and supplementary electronic components e.g. power electronic- and CAN-buscomponents. The total volume of the capacitor modules was 160 liters.
Figure SC3: The fully assembled super capacitor modules.
10
A low ESR of the capacitor modules is a prerequisite for good power performance and high efficiency. The series resistance ESR of both modules in series was 112 mΩ. In order to minimize interface corrosion effects between the electrical contact and the capacitor, aluminum bars were chosen as electrical connectors. A supplementary active voltage balancing electronics was mounted to equilibrate the cell voltages inside the capacitor modules. Unbalanced voltage across the capacitor modules would result in overcharge of some cells and would lead to increased degradation of the capacitor cells and eventually to a failure of the capacitor modules. The energy content and the power of the module, was measured on a dynamic test bench with constant power between full (360 V) and half (180 V) rated voltage. The module was capable of providing a constant power of 50 kW during 15 seconds of discharge, as shown in Figure 10. This is equivalent to an energy content of 210 Wh @ 50 kW. The theoretical efficiency of this discharge process with the above-mentioned ESR is 92 %.
Figure 10: Measured current, voltage and calculated terminal power during a constant 50 kW discharge of the capacitor modules.
11
5 Power Electronics The simultaneous use of two different DC electrical power sources in the powertrain requires the application of DC-DC converters to properly control the electrical power flows [5]. The DC-DC converters are the interface between the drive inverter, the fuel cell system and the super capacitor-modules. Initiated by the drivers demand, the drive inverter transmits the required power pd to speed the vehicle. This power can be delivered from the fuel cell or from the super capacitor or from both systems together. The DC-DC converters allow sharing the power flow of the drive inverter between the super capacitor and the fuel cell in accordance to a reference value calculated from the strategy-controller. For the hardware realisation triple interleaved DC-DC converters were chosen for the fuel cells and the super capacitors. A brake chopper was also installed in the DC-link to reduce its voltage in case of severe failure. The semiconductors are 600V-IGBTs with anti-parallel diodes. The chosen topology is shown in Figure 11.
L
C
L
R
uc
u sc SC
FC ufc DC
ASM AC
inverter
Figure 11: Topology of the DC-DC converters.
6
Results
The powertrain was tested on a dynamic test bench, and implemented in the car on the road. Typical results of the tests on the dynamic test bench are shown in Figure 12. Driving the first 6 km on the beginning slope of the Simplon pass is simulated. In Figure 12 the total power available for the driving motor, the net power from the fuel cell and the power balancing by the super capacitor are shown. For the first 50 s (800 – 850) super capacitor is charged (no motion of car), then (850 – 930 s) very variable load is required during the urban
12
section of the road. Negative power for the super capacitor is charging from the fuel cell. After 930 s, the road is sloping up and the fuel cell is operated at constant power of ca. 30 kW (net). Decelerations and accelerations are balanced by the super capacitor. At 1100s the car stops at a red light and the super capacitor is recharged again.
60 Fuel Cell net
50 Total Power
Power [kW]
40 30 20 10 0 -10 -20
Super Capacitor
-30 800
850
900
950
1000
1050
1100
1150
Time [s]
Figure 12: Test of the complete powertrain on the dynamic test stand. Power flows for super capacitor, fuel cell (net power) and total power are shown when driving the first 6 km of the Simplon pass. Explanations see text above.
13
7
Conclusions
The following main conclusions can be drawn from the design, development, set-up and testing of the fuel cell supercap hybrid electric power train : • • • •
The hybrid powertrain allows for a comparably low power fuel cell system conserving high peak power for acceleration. Due to the hybridization with the super capacitor, the powertrain shows excellent dynamics. Super capacitors have the potential for short-term peak leveling applications. The recuperation of braking energy is an ideal application for super capacitors.
8 Acknowledgement Financial support by the Swiss Federal Energy Office and by AMAG Schweiz AG is gratefully acknowledged.
9 Literature [1] M. Ruge, F.N. Büchi, Bipolar Elements for PE Fuel Cell Stacks Based on the Mould to Size Process of Carbon/Polymer Mixtures, Proceedings of the 1st European PEFC Forum, pp. 299, (2001) [2] M. Ruge, and F. N. Büchi, PE Fuel Cells: Evaluation of Concepts for a Bipolar Plate Design and Construction, Proceedings of the Energy and Electrochemical Processes for a Cleaner Environment of the 200th Meeting of the Electrochemical Soc., PV 2001-23, 165-173 (2001) [3] S. Pischinger, C. Schönfelder, W. Bornscheuer, H. Kindl, A. Wiartalla, Integrated Air Supply and Humidification Concepts for Fuel Cell Systems, SAE Paper 2001-01-0233, SAE 2001 Congress, Detroit /Michigan, (2001) [4] R. Kötz and M. Carlen, Principles and Applications of Electrochemical Capacitors, Electrochimica Acta, 45, 2483-2498 (2000) [5] H. Stemmler, O. Garcia, A simple 6-way DC-DC converter for power flow control in an electric vehicle with fuel cells and super capacitor, EVS-16, 13.-16. 10. 1999, Peking, China (1999)
14
