Innovative Technologies For PublicElectric Transport Systems

Urban Transport 569 Innovative technologies for public electric transport systems G. Martinelli, A. Morini & A. Tortella Department of Electrical Engineering, Padova University, Italy Abstract This paper presents the development status of the innovative technologies adopted for urban electric transport systems. The description is mainly focused on the zero emission buses based on fuel cells and energy storage devices such as supercapacitors and batteries. These vehicles are characterised by higher operating range and efficiency than hybrid and battery buses and don’t need an external supply. The features of the on-board supply system, of the control strategy and of the drive configuration are presented with reference to the guidelines developed in a national research project involving four Italian universities. A detailed analysis is given about the choice of the type of electric motor, as it directly affects the vehicle performance and the drivetrain efficiency; in particular, three in-wheel permanent magnet motors are compared in terms of installation requirements, torque production and total weight. Keywords: electric transport systems, fuel cell vehicles, energy storage devices, in-wheel motors, permanent magnet motors. 1 Introduction The increase of the air and noise pollution which has nearly reached and often exceeded the safety limits also in the mid-size cities requires the adoption of new technological solutions for the urban transport systems with low environmental impact. The need to reduce the traffic of private cars calls for the realisation of innovative public transport systems which must ensure a good level of comfort and efficiency with very low emissions and reduced noise and vibrations. The introduction of new fuels (biodiesel, demineralised water emulsions, methane etc.) as partial or complete replacements of diesel fuel for buses propelled by only an internal combustion engine (ICE), leads to a limited decrease of the WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) 570 Urban Transport dangerous emissions, but it isn’t a key solution for pollution problems. More important improvements can be achieved by using hybrid electric buses (HEV) with series configuration, which is well-suited for urban routes consisting of frequent accelerations/decelerations and stops: for such vehicles the ICE operates at the maximum efficiency point, while the electric motor provides the traction power and recharges the energy storage system during braking (Husain [1]). However only a pure electric vehicle (ZEV) can fulfil the zero emission target. The ZEV currently in operation are supplied by an external overhead wire (tramway, trolley-bus, intermediate rubber tyred systems with guided running mode Andriollo et al. [2]) or by on-board batteries. The latter have relevant limitations as regard the operating range and the required space: due to the very limited passenger capacity, the battery vehicles are not able to replace the existent public systems but they can be applied to particular services like inside the downtown area forbidden to private cars or inside hospital areas or for the connection between interchange parkings. The utilisation of ZEV buses supplied by fuel cells (FC buses) seems to be the most promising solution as regard both the operating range and the efficiency, in case by supporting the FC with onboard energy storage devices. In the present paper the status of the research and experimentations on FC buses is described with reference to an Italian research project (PRIN 2002) financed by the Italian Ministry of University and Research (MIUR) involving four universities. In particular vehicles supplied by energy storage devices like batteries and supercapacitors and propelled by permanent magnet synchronous motors directly coupled to the wheels are considered. 2 Supply by fuel cells and energy storage devices The FC is essentially an electrochemical generator where the reaction between a fuel (typically hydrogen) and an oxidant (oxygen or air) gives electric energy, water and heat as products. Like a conventional battery, the fuel cell consists of two electrodes (cathode and anode) and an electrolyte which allows the ion migration; unlike a conventional battery, the active material can be renewed and so the electric energy is produced indefinitely from a theoretical point of view if the fuel and the oxidant gas are continuously supplied [3]. The FC main parameters cells considered for both fixed installations and traction applications are shown in Tab.1. The operating temperature and commercial cost per kW indicate that the MCFC and SOFC types are hardly applicable for electric vehicles, while they are suitable for stationary energy production, since the efficiency can be also very good. The DMFC type would be favourable as it doesn’t require either on-board hydrogen storage or reforming processes, but the power density and lifecycle are low. Among the other types, the SPFC (also known as PEMFC) has found an intensive development and application thanks to the high power density, the lifecycle and the limited cost per kW: such aspects suggest the SPFC to be the best candidate to supply commercial vehicles. As the output voltage of the individual cells ranges between 0.5 and 1 V depending on adopted technology, a series connection is applied (FC stack) in order to obtain the DC bus required voltage. WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) Urban Transport Table 1: FC type PAFC AFC MCFC SOFC SPFC DMFC 571 Fuel cell main parameters [4,5]. Electrical system Projected life Projected cost Working Power [103 h] [USD/kW] temperature [°] density [W/cm2] efficiency [% LHV] 150-210 0.2-0.25 40-45 40 1000 60-100 0.2-0.3 50-60 10 200 600-700 0.1-0.2 43-55 40 1000 900-1000 0.24-0.3 43-55 40 1500 50-100 0.35-0.6 32-40 40 200 50-100 0.04-0.26 >50 10 200 PAFC: Phosphoric acid fuel cell; AFC: Alkaline fuel cell; MCFC: Molten carbonate fuel cell; SOFC: Solid oxide fuel cell; SPFC: Solid polymer fuel cell known as proton exchange membrane fuel cell (PEMFC); DMFC: Direct methanol fuel cell. The typical operation of FC buses characterised by fixed routes and periodic stops at the depots minimises the requirements of hydrogen supply infrastructure: hydrogen can be directly provided by storage banks or it can be produced by hydrocarbon reforming (for instance methane) by suitable plants located in the depot. The research projects about the utilisation of FC buses have reached an advanced level of experimentation and pre-commercialisation: among the current projects (ECTOS, STEP [6], California’s Fuel Cell Bus program [7]) the most important one is CUTE (Clean Urban Transport for Europe) [6], cofinanced by the European Union, which aims to analyse the data related to reliability, costs, efficiency and environmental benefits obtained from the operation of 27 FC buses in 9 European cities with different characteristics. Different plants for hydrogen production, delivery and storage are also compared. At present the commercial use of FC buses is penalised by the high cost of FC modules and the difficulties related to hydrogen supply and storage. However a significant decrease of such problems is expected in the near future, due to the growing interest and investments for hydrogen production. Furthermore the technological development and a large scale application in other industrial sectors makes probable a remarkable reduction of the FC cost which could equate the hybrid one within ten years. In order to improve the vehicle performance, the FC are often supported by an energy storage system: the rating of the FC module is defined by the average value of the drive power demand, while the energy storage devices deliver the peak power requested during starting and acceleration phases. This solution leads to the following advantages: • reduction of the FC ratings and then of the capital costs; • improvement of the global efficiency of the propulsion system due to the optimisation of the drive supply with consequent reduction of the operating costs especially for routes with high traffic intensity; • improvement of the system performances as the energy storage system is more suitable to compensate the high-rate power demand with respect to the FC supply as the time constant of the hydrogen flow is about ten seconds. The components used for the energy storage can be electrochemical batteries, double-layer capacitors (supercapacitors) or flywheels. The batteries have high energy density and can therefore integrate the average power supplied by the FC. WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) 572 Urban Transport However, some capability to handle short power peaks is often requested: to such purpose the lithium technology (Lithium Ion, Lithium Polymer) seems to be the most promising in the mid-term, thanks to very high lifecycle and good strength to deep charge and discharge processes. Unlike the conventional electrolytic capacitors, the supercapacitors can reach capacity of thousands Farads in very small volumes, though the cell operating voltage is lower than 2.5 V. For the power levels needed for urban buses, the cells must be series-and parallel-connected (supercapacitor banks) and an auxiliary circuit must control the static and dynamic distribution of the cell voltage. The supercapacitors favourable features are: • higher power density than batteries; • fast charge and discharge processes (time duration of few seconds); • high number of cycles (order of hundreds of thousand) without shorten the component lifecycle, as otherwise occurs for the batteries; • reduced maintenance; • lack of toxic materials in the cell composition. However the supercapacitors have an energy density much lower than the batteries (as well ten times lower). As concern the energy storage system of a FC bus, the most favourable solution seems to be a combined utilisation of batteries and supercapacitors (Andriollo et al [8]). The batteries provide high energy density, while the supercapacitors provide high power density by ensuring fast and repetitive charge and discharge. The main benefits are: • increase of the energy storage efficiency and reliability; • increase of the vehicle performances and operating range; • longer battery lifecycle; • reduction of the vehicle weight as part of the battery pack is replaced by supercapacitor banks. The definition of the design guidelines and the performance analysis of FC buses are the subjects of an Italian research project which involves the Universities of Cassino, Naples, Padua and the Polytechnic of Milan: the main research items concern the adoption of both batteries and supercapacitors as energy storage devices and permanent magnet synchronous motors as traction drive. The layout of supply and traction system is shown in Fig.1. Two solutions are considered for the traction motor: in the former (A) the motor is coupled to the axle by the gearbox and the differential (conventional solution), in the latter (B) the motor is placed inside the wheel (in-wheel motor) and it is mechanically coupled to the rim without the gearbox interposition. The motor is supplied by an inverter connected to the DC bus which collects the power from the FC, the batteries and the supercapacitors. Such devices are parallel-connected to the DC bus usually by a bi-directional (batteries and supercapacitors) or unidirectional DC/DC converters. The control system manages the energy flow related to each source by generating the drive signals for the DC/DC converters, the inverter and the system regulating the hydrogen flow. A typical progress of the power WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) Urban Transport 573 exchange between the energy sources is represented in Fig.2, related to the different operating conditions of a standard drive cycle. Wheel Traction unit AC Motor Trasmission + Differential (A) Wheel Inverter Wheel AC Motor (B) DC Bus Fuel Cell Supercapacitors Battery Control unit Figure 1: On-board electric supply system of a FC bus; (A): traction unit with AC motor and gearbox; (B): traction unit with AC in-wheel motor. Energy supplied by batteries and supercapacitors Power Recovered energy for the recharging of batteries and supercapacitors Mean power supplied by the fuel cells A B Figure 2: C D E F G Time Power flow during the vehicle operation. As can be noticed from the diagram, the mean power required by the traction drive is provided by the FC, while the amount of power exceeding the mean value can be drawn from the supercapacitors and the batteries (intervals A,B,E,F): in particular during the acceleration phase A it is convenient the utilisation of the supercapacitors as the vehicle demands a relevant and concentrated amount of power. On the contrary when the power demand is lower than the mean value, the FC recharges the energy storage devices (intervals WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) 574 Urban Transport C,D,G): in particular during the braking interval G, the control unit directs the recovered energy mainly to the supercapacitors for which faster recharging time than batteries can be achieved. 3 Electric motors for the urban propulsion At present the electric motor more frequently used for ZEV propulsion is the induction squirrel cage one. The motor is coupled to the gearbox and located in the front or rear of the vehicle or near the wheels connected to the rim. The main advantages of the induction motor are the easy construction, the high mechanical power delivered to the axle, the low cost per kW and the high operating reliability. However it is unfavourably affected by the low global efficiency due to the transmission losses related to the gearbox which also requires frequent maintenance. In addition the relevant weight and volume of the motor-gearbox assembly results in a further reduction of the available room which is already restricted by the presence of the FC and the energy storage system. In order to overcome the previously described drawbacks, an alternative solution proposed for some experimental buses consists of the adoption of brushless permanent magnet motors. The main features of such motors are: • high specific power and power density with reduction of weight and volume; • high efficiency and power factor due to the absence of the magnetizing current which must be provided to the induction motor; • large air-gap width which enables a better fit of the mechanical tolerances; • possibility to use magnetic configuration with high pole number: this allows to couple the rotor directly to the wheel (in-wheel motor), with improvement of the total efficiency because of the gearbox removal and reduction of the motor weight because of the lower flux per pole; • use of lumped coils easy to be manufactured: their shape allows a remarkable reduction of the motor volume due to the short end-windings and an easier installation of the stator water cooling system. The disadvantages are higher manufacturing costs, the possible demagnetization of the permanent magnets at high temperatures or in short circuit conditions and the reliability, not so tested as for the induction motor. The studies carried out in the research project are mainly focused on the analysis of in-wheel motors, with the outer rotor connected to the rim: this solution is advantageous because it is very compact and reduces the moving parts, as the axle is fixed. The permanent magnet motor may be realized according to different magnetic configurations and the choice depends also on the mechanical and electromagnetic constraints imposed by the kind of installation. In order to reduce the rotor manufacturing and assembling difficulties, only configurations with surface magnets are considered, with the permanent magnets located on a solid core, which may be easily and strongly fixed to the rim. Three types of motor are compared, differing in the way in which the flux is distributed in the magnetic circuit: the radial flux motor (RFPM), the axial flux motor (AFPM) and the transverse flux motor (TFPM). The electromagnetic configurations of the motors and their assembling schemes inside the wheel are shown in Figs.3–5. WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) Urban Transport Tyre (a) Permanent magnets (b) 575 Rotor core Rim Coil Magnets Bearings Cooling Stator Fixed shaft RFPM Figure 3: (a) Phase coils In-wheel motor with RFPM configuration; (a): motor assembly inside the wheel; (b) 3D view of the magnetic configuration (two polar pitches). (b) Tyre Stator Phase coils Rim Magnets Stator Stator Rotor core Coil Bearings Cooling Fixed shaft AFPM Figure 4: (a) In-wheel motor with AFPM configuration; (a): motor assembly inside the wheel; (b) 3D view of the magnetic configuration (two polar pitches). Tyre Rim Magnets Permanent magnets Rotor core (b) Rotor core Stator ‘C’ cores Permanent magnets Stator Coil Bearings TFPM Figure 5: Cooling Fixed shaft Phase coils In-wheel motor with RFPM configuration; (a): motor assembly inside the wheel; (b) 3D view of the magnetic configuration (two polar pitches). It must be pointed out that the assembling of the wheel on the fixed axle must be made in order to have very small vertical or lateral clearances, otherwise the consequent motor air-gap variations would produce very high attraction forces between the permanent magnets and the stator cores. WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) 576 Urban Transport The use of a liquid cooling system is supposed for all the motor configurations in order to get the torque values needed for the normal vehicle operation: in starting and sloping conditions the needed torque may be more than 1000 Nm, if only the rear wheels are supposed to be motorized. The main characteristics of the RFPM configuration are: • easy manufacturing, because well-known technologies are used; • possibility to improve the motor performances by increasing the motor depth (the diameter being equal); • reduction of the torque ripple by using a suitable shape for the stator teeth or by skewing the magnets (with detriment to the construction ease and to the motor cost). In the AFPM and TFPM configurations the torque is exerted on two sides of the rotor surface (double-side configurations): this solution reduces the effects in the vertical direction of vibrations and accidental shocks and makes possible the torque ripple compensation by displacing the magnets on the two sides of the rotor (see Fig.6a for the TFPM motor). However, the phase coils are split to reduce the volume and facilitate the installation of the stator core, making the winding assembling more difficult, in particular in motors with several poles. On the other side, these configurations make possible to connect the coils in series or in parallel, to adjust the voltage and current to the values of the supply system or to compensate the torque decrease at high speed. With reference to the AFPM configuration, the stator weight may be strongly reduced by removing the ferromagnetic yoke, because it is useless for the closure of the magnet flux; anyway, this would make more difficult to join the stator with the axle. The TFPM configuration exhibits a very light rotor, because the only function of the rotor core is to support the magnets; anyway, its vertical position makes more difficult the mechanical connection with the wheel rim. Such aspect makes the assembling quite difficult, because the rotor must be inserted between the stator cores already set up on the shaft; moreover the higher size in the radial direction reduces the room for both the coils and the cores, with consequent increase of the magnetic saturation and reduction of the total ampereturns. Among the favourable features it’s worth to point out the good cooling efficiency with the cooling system positioned also between the coils of the same core (Fig.6b) and the possibility to increase the torque by using two rotor modules, the radial size of the motor being maintained (Fig.6c). The comparison between the performances of the motors in terms of developed torque is carried out by considering the same permanent magnet weight and the same maximum current density: the first quantity mainly determines the motor cost, the second defines the maximum thermal load congruent with the insulation capability and the performance reduction of the permanent magnet. In addition, some geometrical and electromagnetic design constraints are considered; among them the maximum motor radius (Rmax=287 mm), the air-gap width (g=3 mm), the maximum flux density in the magnetic materials (Bmax=2 T) and the minimum flux density in the magnets to avoid demagnetization (Bmin=0.35 T) (Andriollo et al [9,10]). WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) Urban Transport (b) (a) Outer cooling Rotor (c) Inner cooling 577 PM flux ∆ Stator Features of the TFPM configuration; (a) magnet displacement ∆ to compensate the torque ripple; (b): positioning of the cooling system; (c) configuration with two rotor modules. Figure 6: 1400 1200 [Nm] 1000 800 600 400 RFPM AFPM TFPM 200 [A] 0 0 25 Figure 7: RFPM AFPM TFPM 75 100 125 150 175 Torque as function of the supply current for three in-wheel motors. Table 2: Motor type 50 Weight and depth of in-wheel permanent magnet motors. Copper weight [kg] 15.0 24.0 21.0 Rotor weight [kg] 28.0 26.0 13.5 Stator weight [kg] 47.0 58.0 61.5 Total weight [kg] 75.0 84.0 75.0 Depth [mm] 110.0 162.0 170.0 The developed torques as function of the supply current are compared in Fig.7; the curves are related to 24-pole motors and calculated by means of a finite element code, considering sinusoidal operation (frequency 102 Hz - that is vehicle speed 60 km/h - and maximum current density 10 A/mm2). The RFPM and AFPM curves are both practically linear with torque values reaching 1300 Nm (motor maximum power 70 kW), while the TFPM motor is more affected by the magnetic saturation and the developed maximum torque is about 15% less. The comparison between the weight and the depth of the motors is given in Tab.2; the data show that the TFPM configuration makes possible to halve the weight of the moving part, to the detriment of the transverse size, which anyway does not exceed the wheel cross-section; the AFPM configuration is the heaviest, even if the stator weight may be reduced (about 10%) by removing the yoke without torque reduction. In conclusion, the RFPM configuration is the most promising in terms of performances, sizes and construction ease. WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line) 578 Urban Transport 4 Conclusions The advanced status of the research programs and experimentations on the FC buses proves that such vehicles have nearly reached performances and operating reliability which makes probable their introduction to commercial operation in the mid-term. As pointed out also in the studies developed during the Italian research project involving the Universities of Cassino, Naples, Padua and the Polytechnic of Milan, the most favourable solution for the urban operation consists of the adoption of the FC with an energy storage system which combines batteries and supercapacitors: this solution allows to achieve a significant reduction of the FC ratings, a reduction of the hydrogen consumption and an increase of the lifecycle of power devices, mostly of the batteries. Moreover the utilisation of in-wheel permanent magnet motors can lead to a further improvement of the global efficiency and the operating reliability due to the gearbox removal as well as a reduction of the weight and the volume of the traction unit. A comparison among different magnetic configurations with surface permanent magnets shows that the radial flux motor is the most convenient for both the construction ease and the high torque to weight values. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] Husain, I., Electric and Hybrid Vehicles – Design Fundamentals, CRC Press: pp.243-260, 2003. Andriollo, M., Del Torchio, G., Martinelli, G., Morini, A., Tortella, A., Innovative technologies for the public transportation in the Urban Mobility Plan of Padova, Urban Transport VIII - Urban Transport and the Environment in the 21st Century, WIT Press, pp. 75-88, March 2002. U.S.Dept. of Energy (DOE), Fuel cells technology handbook, EG&G Services Parsons, Inc.: pp.1-35, October 2000. Chan, C.C., The State of the Art of Electric and Hybrid Vehicles, Proceedings of the IEEE, Vol.90, no.2, February 2002, pp.247-275. Apollo Energy Systems, www.electricauto.com/_pdfs/fc_chart2.pdf. CUTE, ECTOS, STEP Projects, www.fuel-cell-bus-club.com. North-east Advanced Vehicle Consortium, www.navc.org/California .html. Andriollo, M., Castelli Dezza, F., Tamburrino, M., Converter Control Strategies for the Power Flow Management in a Fuel-cell Supplied Vehicle for the Public City Transportation, Proc. of SPEEDAM 2004, Capri, Italy, 16-18 June 2004. Andriollo, M., Martinelli, G., Morini, A., Tortella, A., Zerbetto, M., A Transverse Flux Wheel Hub Motor for Electric Buses, Proc. of SPEEDAM 2004, Capri, Italy, 16-18 June 2004. Andriollo, M., Forzan, M., Morini, A., Martinelli, G., Tortella, A., Zerbetto, M., Performance Analysis of a Transverse Flux Wheel Motor by a Non-linear Mathematical Model, Proc. of ICEM 2004, Krakow, Poland, September 2004. WIT Transactions on The Built Environment, Vol 77, © 2005 WIT Press www.witpress.com, ISSN 1743-3509 (on-line)